Loading...

Follow Straight Path Solutions - SQL Server Consulting on Feedspot

Continue with Google
Continue with Facebook
or

Valid

Hello!

Thanks for attending! The slides for my SQL Saturday Session on High Availability and Disaster Recovery are below. After the session I’ll update this with links and notes from the talk by end of day Tuesday. Thanks for attending!

The post SQL Saturday 845 – Session Details appeared first on SQL Server Consulting - Straight Path Solutions.

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

This article is part of our SQL Server Upgrades Made Easy series:

  • Why Upgrade?
  • Before You Upgrade
  • During Your Upgrade
  • After Your Upgrade
  • Recap Chat
Upgrading SQL Server – Week 4 “After You Upgrade SQL Server”

You can watch the previous 3 weeks on youtube

This is the last video from the webinar portion of the SQL Server upgrade webinar series from the past four weeks. Here we chat about the things to consider AFTER your SQL Server upgrade.  Health checks. Monitoring. Keeping the lights on. Frankly, this webinar is a must whether you are doing a SQL Server upgrade or not! In a week – after our live hangout video goes live, we’ll make a single page that has all of the videos, download links, references to blog posts and other content you might like and if we come up with more free checklists or downloads to help you with your SQL Server upgrade projects, we’ll throw it there and let that page be a living page. Subscribe to the blog or come back in a week to check it out!

After Your SQL Server Upgrade - Upgrading SQL Server Webinar Episode 4 - YouTube

You can also sign up with the webinar ninja series to get the full cuts from Webinar Ninja series and watch replays there and see any Q&A each week including the little crosstalk before and after the “recorded” portion of the webinar starts. Might even see me “dancing” sadly!!

Subscribe to the youtube channel to be alerted of the live chat hangouts and to see the upcoming talks in our continuing webinar series.

Towards the end of March, I’ll start with a multi-part series on High Availability/Disaster Recovery options for SQL Server. If you have questions, you may not need consulting, just some help!  That’s presicely the point of these webinars. Trying to help get some content to you of things to think about when there just isn’t a training budget but you’d like level up a bit. Perhaps in the future we’ll go 200/300/400 level and consider freemium and premium content. For now, though, these webinars will be live, recorded, and free to watch either way.

I hope to see you in the Google Hangout with some of the Straight Path team and I discussing upgrades on the live chat thinga-ma-jiggy at this link.

Downloads from the SQL Server Upgrade Series

(We’ll have a couple more pieces of content still to come on the final recap post that links to all the videos, and all the downloads and plus we’ll work to get links from all the videos when we make that “Central Page” for the webinar series! The upgrade checklist is ready but we’ll put that into it’s own post today, 3/7/2019!)

Transcript

(I’ve used a transcript service for this and I reviewed some of it. It’s a work in progress and there may be some errors!)

Welcome back to the webinar series. Today’s the last actual webinar in the upgrade series, so we’re gonna talk today about things to consider after you upgrade SQL Server. Again, like I’ve said a couple times in the blog post, and like I mentioned ahead of time, this webinar really has little to do with the SQL Server upgrade, but it’s my chance to extol some of my virtues on you when it comes to best practices and configuration and desired state and all the stuff that we should be worrying about but sometimes don’t have time to do. So, like we talked about before, when you’re done with a SQL Server upgrade or when you’re doing a SQL Server upgrade, this may be something that a company might do once every five years, once every ten years as some people here would say, and if we’re doing it so infrequently, what better thing to do than to use this as an opportunity to start fresh.

Pretend it’s like New Year’s Day for your SQL Server. You’re gonna finally eat better and workout or whatever your resolution is, except for this is gonna be much more serious than New Year’s resolution because you’re actually gonna do it, so when you’re done, how do we know we have a best practice to set up on our SQL Server? How do we know our SQL Servers are set up for success? And, that’s what we’re going to talk about today.

If you’re new to the series, this is where we have been. We’ve done a lot of things. We had a big picture at the very beginning. That was more just talking about why, why, why, so why are we going to upgrade? What’s the point?

Then, we talked about what do you do before you upgrade? How do you plan it? How do you ask the right questions and get the right people involved? We talked about checklists, and that’s one download I’m still working on finalizing for you, so that’ll be in the recap. We do have one PDF we’ll talk about in a second, to download today. From last week, the SQL Server health and best practice checklist, we finally released that, so you could download that on a blog post that went up today.

Then today, we talked last week about the upgrade, so how do you actually upgrade? How do you actually go in there and do your upgrade, we talked about … Again, in 30 minutes, I’m not gonna tell you how to upgrade your SQL Server from start to finish, but we talked about the important concepts.

I think the HA/DR web series might be a little more involved. We can stretch it out a bit and maybe do a few more weeks in there and actually do some demos, and AGs and FCIs and even log shipping. But this, I don’t wanna get into demos because there are so many. I’ll post the link. Actually, I’ll post it right now. If you just go to my blog. Just go to my blog, Jose, and you’ll see the latest post is the recap from week three.

In there, there’s two links. There’s only two downloads right now. There’s the 20-ish page SQL Server Configuration Guide, and then there’s the much smaller PDF of the Do Your Own Health Check guide. That’s cool, but the actual SQL Server Configuration checklist is a recap of all of our best practices that we look through when we do a SQL Server upgrade. Checklist is probably a strong word for it. It’s really a document with bullet points, but I’m sure we can probably turn it into a checklist if people would like it that way.

Then, the last thing I’m working on is an actual upgrade checklist. Again, I proposed … I might just make an Excel spreadsheet and post that instead of trying to make it PDF and make it marketing friendly. That’s the hang up here, is trying to get into PDF, but I’ll figure out the right format, and I’ll put that, when I share I this video from this series next week, I’ll share that.

Next week, I’m not traveling as much, so I should actually have time to get the blog post out by probably Monday or Tuesday, like I hope, and get the transcript servers to run sooner. Anyway, that’ll be the last downloadable thing we have there, is the upgrade checklist, or the before and after and steps there.

Anyway, we’re here today, we’re going to talk about after. We’re going to talk about what do we consider when we’re all done upgrading, you’re not just done when your upgrade is done. And if you think you’re done when your upgrade is done, you’re lying to yourself. You should never be done with SQL Server. We’ll talk about that in a second.

Then next week, we’re going to do a live hangout. So, this will not be on WebinarNinja. You have to go to that link right there, that YouTube.com watch and that will take you to, I think, to my channel, the Straight Path, our channel I should say, the Straight Path channel. Once you’re there, there’s a little remind me or whatever icon, you click that icon and Google will somehow let you know, maybe give you an iCal to remind you for the live hangout. Next week, same time, we’ll have the whole Straight Path technical team, or a good portion of us anyway. I know Tara will be there. I know I’ll be there. I think Mike Clark will be there, I’m not sure. And maybe Jack and Joe, and maybe Bruce. So, we’ll have the whole technical team there to answer your questions.

If you have questions, we’ve gotten a couple, if you have more questions, whatever they are, if it’s about a Sequel upgrade or even about the health checks and the best practice stuff, bring the question with you live. Or if you’re too shy or you can’t be live, we’ll record it, just send me an email to mike@straightpathsequel.com and we will answer that question on that hangout. We’ll record it, and when this whole series is all done next week, when I do that hangout, I’ll take the video from that and all these other videos and blog posts, and we’ll make one big, sort of like a spoke or hub blog post that lists all the videos in order with more download links. And as I re-watch the videos and listen to my annoying voice, I say, “Oh, I should probably add that,” so I’ll add a lot more extra links that nobody asked about but that come to mind while I watch it. So, that last post will hopefully be a guide and maybe even have some checklist steps in there.

The goal again, is not to upgrade your SQL Servers in five 30-minute webinars. It’s really to get you thinking about the right things, and to help get you your questions answered and to put you on the right track.

So, thank you for coming back, by the way. A lot of you came back and I appreciate that. You stuck with us, so either you have nothing else to do or you appreciate the content.

Who am I? You all know by now. Most of you know who I am. I’m Mike Walsh. I run Straight Path Solutions. We’re a SQL Server consultancy. I’m not doing this as a SQL Server consultancy thing, I’m doing this really as an MVP thing. I’m just doing this as a person in the community trying to give back. Whether I keep my MVP or don’t, I don’t care, I just like sharing. Our company does a whole bunch of things. But one of the things we do a lot of, being a consultant company, is come alongside people and do upgrades.

As a result, I’ve been burned by a lot of upgrades, and I’ve taken over upgrades that they didn’t think about these things ahead of time. They didn’t go through a checklist. They didn’t do best practices. We’ve been there to catch people when they fall from upgrades not well thought out. I want to avoid that for you. I want you to avoid that.

Where are we? Again, there’s some recaps if you go to straightpathsequel.com/blog you’ll see pretty much all the latest posts there. There’s one availability group post in there. But, the other latest posts are all about upgrades. I have a couple more posts in the queue, in the draft mode, that will be about upgrades as well, and again, we’ll have the big recap post next week with a bunch more links. Really, the content for the rest of the next week or two on my blog will be about upgrades. I’m going to switch over and make it HA/DR stuff, because that will be the next webinar in March.

Again, the same idea. We’re not going to go in depth. We’re not going to be 500 level material. I want to start at the beginning and what should you think about before. Which approach, how do you do RPO and RTOs, and how do you balance budget, and can AGs do this? Can FCI do this? How do we mix and match, and how can we do it on a shoestring and what shouldn’t we do on a shoestring, and all these things. I want to talk about before, during and after. In fact, every single webinar we do, I really want to hammer this whole idea of things to think about and do before, things to think about and do during, things to think about and do after. I think if we simplify it to that, we can make simple. Right? That’s my hope.

The next series will be probably two weeks after the live hangout, so towards the end of March. That will be on, again on High Ability DR, so the content will start switching over the HA/DR at that point, I think.

By now, we should know all the things we knew last week, the week before. We should know why we want to upgrade. We should know what we need to think about. What questions we should be asking. You know, who, what, when, where, and why, from two weeks ago video. We should really be knowing where we want to go, how we want to get there, and really why we want to go.

Now, we should also know how because last week we talked about how. If you all remember one thing from last week, I don’t like in place upgrades, and really frankly, you shouldn’t either. It doesn’t save you time and the rollback can be bad. If you have no choice, you have to be careful and plan it out, and spend extra time making that rollback checklist in the go, no go criteria. But you should have a better idea of how you want to upgrade. How you’ll test it. Who’s in charge. That sounds so simple but so many times you can have conversations and somebody thinks that they just got assigned a task … or they think they assigned you a task but nobody knows because nobody’s just done that little … here’s the plan, here’s your role, here’s your role, he’s your role, and people who asked before, I forget who it was now, maybe Dale, you know, what if four weeks ago, you have all the hats? You’re the DBA, the SysAdmin, or again, well have the conversation with yourself again to understand what the right roles are and what the right pieces are.

Then, we’ll talk about the rollback. Talk about the steps, and that’s what we should know from the before, and during upgrades. That’s it. We’re upgraded. Congratulations. The webinar series is over, right? No. No. Remember we talked last week about this. Y’all remember what this is, right? You don’t have to all type in chat but it takes hard work to type a few keys, but we know what this is. This is the Show Time rotisserie grill. Actually, I think my in-laws have one and it actually works okay. You can cook Italian sausages in it, by the way. But I’m not [Roan Po-peel 00:10:07]. It’s really good if you don’t like greasy Italian sausages in spaghetti. But anyway, I really digress now.

This is a chicken making machine. It’s set it and forget it, is what he had all the people in the infomercial audience saying over and over. “Set it and forget it.” This is not the SQL Server rotisserie chicken machine. This slide right here, while I’ve the changed the SQL Server logo, the very first talk I ever gave on SQL Server, one of my slides was the rotisserie chicken machine with a SQL Server 2008 logo on it. Probably 11 years ago. I had this presentation I gave called, “So You Want to be a DBA? Where Do You Start?”

So many people, not you because you’re here at the webinar and you care about the Sequel community and you care about learning, but so many people think that SQL Server can be set and forgotten. Right? You install it really easy. There’s a few more clicks to have to do, but for the most part, it’s next, next, next, next, finish. Then you can go away, have some of your, no brand name, have some of your coffee, a little sugar. It’s Americano. But you can just walk away and come back and poof, you have a SQL Server. Then, you can take your installer from your client, for your ERP system, or whatever application you’re using to do critical business, and you can install that application and it will create it’s own database, for the most part, or maybe you have to create it. Then you can walk away and you have a SQL Server.

I can’t tell you how many health checks I do for clients, big and small clients, where it’s like, “Oh, you don’t take backups? Did you know you don’t take backups?” “You don’t do check DBs. You don’t do this.” I see it all the time. Not you. None of you do that. You’re watching this video, your here at this webinar. You’re not doing that because you care about best practices, but other people do.

SQL Server, and I’m not a slide guy so you can make fun of my animations all you want, but there you go. That’s the extent of my animations. Rotisserie chicken, SQL Server, no. A bouncing no, too. And a misshape, but SQL Server’s not set it and forget it. That’ the main point of this one webinar. We’re going to talk about just again, things to keep in mind.

This almost goes back to one of the before webinars where I told you that you have to do a health check first. If you were obedient and you did your health check before, you’re going to be okay, because you saw what was bad and you care, and you made a list, a checklist even, of what you want to do new and better on the new server and you fix it. Right? Didn’t you? I hope you did. I hope you’re going to.

If you didn’t do the health check and you’ve upgraded and you did a set it and forget it upgrade where you just do your upgrade and your done, well, it’s not too late. You can do a health check. In fact, I recommend you taking one of the free health tools out there from all the people in the Sequel community who worked hard to build them, and do a SQL Server health check once a year. You know, does DBA checks, [inaudible 00:13:21] I don’t mention here and in fact I actually, I’m going to make a note here.

Normally, our office manager comes to the meetings to take notes but she’s actually on a client call. Some of us have to do work. So I’m just going to make a note here in my notebook. Give me a second.

I’ll make sure I include some links to other tools. I mentioned Glen’s script. I mentioned Brent’s blitz scripts, but there’s so many other ones. And I’ll mention a few others. Try a different flavor. But, once a quarter, once a half year, once a year, do yourself a favor and give yourself a SQL Server health check.

Why, Mike? The configs don’t change. I know they don’t change but new databases got created. New users got created. New logins got created. All things have changed. Maybe not your max [dop 00:14:01] and your max memory. Hopefully that’s not changing as you go, but it could. You could have other people who are helpful DBAs in there, or vendors with the wrong level of access because you didn’t fix security, undoing your best practices.

So, after you upgrade, and if you never upgrade, and if you go to the cloud you never have to worry about upgrades, do health checks. Make sure your server’s good. Seriously, it saves a lot.

Dale is saying, great point Dale, he says, “Applications change and new store procedures get installed. They may not perform as expected,” and there are some client’s tools out there that will actually change your system configuration because the vendor thinks they know best and they need SA for the installer, and the installer will actually change some settings. It’ll turn XP command shell on, and then when you get audited in nine months, the auditor says, “Why is XP command shell on?” And you say, “It’s not. I turned it off.” “No, it’s on.” And you say, “I don’t know who did it,” and you turn it off and then you get a phone call from accounting saying that the system’s not working.

You can not just install and walk away. The DBA, whether that’s your official title or not, is the person at the company, again I’m being, this is hyperbole, sort of. But you as the DBA, I think it stands for Database Advocate, you are the one who cares about the data. In fact, I’ll say … and you can write me a letter if you like, you’re the only one who cares about the data in the company. That may not be true, but you should act like it’s true. Act like your the only one who cares about the data, and treat it well. That means do health check, that means when that bad application vendor tries to do something naughty, you get mad at them and you say something about it.

Some best practices evolve, right? If you were to get a SQL Server health check from me, 10 years ago, before … I think it was before Glen started writing his diagnostic scripts, maybe about the same time, but anyway, 10 years ago before I had a blitz or Glen’s scripts and I had my own …

I had a blitz subscription. I kinda had my own. I bet you there are things that I would tell you about today, that I didn’t check 10 years ago. There are things that I forget about today, or that I find today, that I didn’t even think of three years ago. So some best practices evolve and it’s okay to learn but the trick is you have to keep learning. Right? So towards the bottom in italics there, I say, “Always learn.”

Part of what you do after your upgrade is don’t just move on to other things and forget about your basic PBA skills. Go to classes, take webinars, come to the free HA/DR webinar. Maybe we’ll do one on SQL best practices and performance tuning and configuration. Learn about SARGable queries.

I know we should all know this but I keep finding the same problems over and over so somewhere, some place, people aren’t learning. And, so, if we learn and we do more, we’re good.

So I’m reading from the bottom up. But do the health check. Again, our PDF is a real basic step. So over the two downloads we have, the PDF. One of them is just a real small, “Don’t forget about these important things and people get burned by these things.” Those are the things we see most often getting burned by. The longer document, the one that came in today that we just kinda- hot off the presses today, that has a lot more details about how we configure SQL.

But the trick is be intentional. Intentionality is important, right? So here’s a list of things. Right? What’s your max memory set? Did you set it? Again, it’s never said. Do you have power saving enabled? That’s always set these days still. Is your Tempdb configured best practices?

The nice thing about 2016 and 2017, by the way, is the installer tries to manhandle you and force you to do Tempdb the right way. It even forces you to turn on, if you click the little box, you understand what the box says,..

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

This article is part of our SQL Server Upgrades Made Easy series:

  • Why Upgrade?
  • Before You Upgrade
  • During Your Upgrade
  • After Your Upgrade
  • Recap Chat
Upgrading SQL Server – Week 3 “During Your Upgrade SQL Server”

You can watch week one here on the blog or directly on youtube. You can also check out week 2 on youtube or blog. Please consider subscribing to the channel as we’ll make it a rule that all webinar content from the free webinar series will be shared on youtube and sometimes here on the blog also.

This is the third video from the series from the webinar last week. Here we chat about the things to consider DURING your SQL Server upgrade.  We discuss questions you should be asking yourself before any SQL upgrade project starts. We talk about rollback plans, CNAMEs, checking health before, etc.

During Your SQL Server Upgrade - Upgrading SQL Server Webinar Episode 3 - YouTube

Sign up for the to get the full cuts from Webinar Ninja with any Q&A each week including the replays of the past webinars. Subscribe to the youtube channel to be alerted of the live chat hangouts and to see the upcoming talks in our continuing webinar series. Towards the end of March, I’ll start with a multi-part series on High Availability/Disaster Recovery options for SQL Server. If you have questions, you may not need consulting, just some help! Check out the series, attend live and ask your questions or come to the Q&As. I’ll try and answer most questions left in the posts also. Yes, I own a consultancy, and if you need in-depth help, I think we can help – and I trust you can figure out how to reach us. But the point of this content and the webinar series is to help encourage you and show you how can you do more than you realize! See you on Thursday 2/28 for Week 4 – “After you upgrade”. Then on 3/7 we’ll do a live hangout on Youtube with the Straight Path technical team answering your questions.

Downloads from the Series

(The upgrade checklist download will come out in a separate blog post by itself next week. And after the Live Hangout on Youtube, we’ll take all the videos, all the content and make a single page with all the content)

Transcript

(I’ve used a transcript service for this and I reviewed most of it. It’s a work in progress and there may still be some errors!)

So today, we’re going to talk about doing the upgrade. Today we’re going to talk about … we know what to do before, we know why. What are the steps we should do to actually have a successful upgrade? And what are some pitfalls we can avoid? I need to get over here because my slides are being controlled differently, apologize. Let’s see if I can do this right. Perfect, I can. So, again, this series, we’re really … Oops, yeah, I did it too fast.

We’re in week three, and here in week three … let me see if I can move my mouse right on top. There we go. Here in week three, we’re talking about doing the upgrade. So, some best practices for the new SQL install. How can you minimize downtime, right? How can we get the environment upgraded with having our users waiting nervously for less time? How can we do it with success? What are some ways to test to know we got all the data?

Next week we’ll talk about after the upgrade. Really, next week has nothing to do with upgrades. It does because we should be looking at these things after we upgrade, but if you don’t upgrade your SQL server and if you’re not planning on an upgrade, next week’s going to be packed with good content for you to just be a better DBA and to watch your environment better. So who am I? I’ve said this each week. My name is Mike. You can find out more about me on my blog, StraightPathSQL.com. I run a company called Straight Path. We do SQL server consulting, manage services, we help people upgrade projects. We do all sorts of SQL work.

I have a lot of upgrade scars. Being a consultant, we get to do a lot of upgrades, right? So when I was a full-time DBA, I could probably count on one or two hands the number of times I had to upgrade a SQL server. The only reason I have two hands is because I was a job hopper early on. I was always a consultant, I just didn’t realize it. Most people just don’t do a lot of upgrades, but we do a lot of them as consultants because we have a lot of different clients at different phases of the life cycle. So we’ve seen the problems, and we’ve gotten called into the upgrades that went, oops, or we’ve had the people who didn’t heed advice and tried to upgrade a path that maybe we don’t like, and got burned by it.

So where are we? Again, I already kind of gave the recap. These are the things we should know though. After the earlier two webinars, we should know why we want to upgrade. Obviously, that’s a pretty easy question, but we should know some of the reasons why we want to and some of the reasons why we can … the reasons that maybe our management wants to hear. So, hey dear leader, here’s why you should want me to upgrade. So, that little reverse psychology. Then we’re going to … We already talked about where we want to go, so we should have an idea of that. Again, you need to do the upgrade analysis with the upgrade advisor. We talked last week about the DEA tool, the database experiment analysis tool. We can talk about doing a SQL server database tools last week and building towards your target. The point is, we should know roughly where we want to be.

We should have an idea of what our environment looks like. We talked last week about running a health check and I have you guys … talked about link, and that link for the download is actually on the blog post from week two’s video already. We can run our health check PDF, or you can go and download the blit scripts, the Glenn Berrys scripts, or any scripts. I don’t care. That’s well trusted by the community, and see where you are. We should have a performance baseline, so we make sure that the new environment we build will have the right performance.

We should have the rough idea of which approaches we want to use to test. How do we know that we can support the new capability mode, because going to the new version is really cool, but it’s really important to also be able to work in the latest compatibility mode if we can so we can take advantage of the new features. We can use [inaudible 00:03:50] functions for example. Then we should have an idea … We talked last week about role back and sort of go/no go criteria. I said, it’s not a go go. It’s just go/no go. Most companies don’t have a no go. They just have a go and we’re going to go whatever you say. But we need to be the pesky annoying DBAs who say, yeah but, and explain the problems. So, that’s kind of there we should.

So where are we going to go? Well, today is going to be about how. So we’re going to zoom in on the how. How should we install SQL? We’re not going to go into in depth on that because that’s a whole webinar by itself, but we have a download we’ll have in the blog post about this for that. Like I said, I can’t give you the links to webinar ninja. The people watching the videos won’t be able to get them anyway, so we’re going to put them on the blog post. What’s going to happen is we’ll do a webinar on Thursday, and then by Tuesday of the next week … so three business days later, there’ll be a blog post with a video and a transcripts and links we talked about, and any downloads that we say we’ll make for you.

Normally this downloadable content, we’ll probably use a lead magnet on our website and people can come and give us their email address and all that. But for people coming to the webinar, I just want it to be a free thing for you, so you’ll just have a direct download link. Get it. I don’t want your email address. Just take it and use it. So we’ll have a checklist that we use to help guide us on SQL sort of best practices. That checklist alone could probably be a three part webinar series, but we’ll give you that.

How should you organize your own checklist for the upgrade? We’re going to talk about that today quite a lot. Then again, next week on the Tuesday, we’ll have a sample checklist that you can use and fill in. We’re going to talk about how to best execute an in place upgrade versus a migration based upgrade. How do we minimize downtime? How do we do a test ahead of time? How do we do a test after? These are the things we’re going to talk about this week.

So let’s first … I just want to talk about in place upgrades first, get this out of the way. I have a pretty simple in place upgrade process. My first step for in place upgrade is you have to attain the SQL server installation media. So go to MSDN, go to your volume licensing portal, wherever you get it. You can even get the old DVD. So this is SQL server 2012. At the [inaudible 00:06:03] they gave us a pretty cool sign book. The DVD’s not in there. It’s just an empty package with all the signatures of the program team.

It’s kind of neat. I keep it up here. It’s an old version of SQL and I don’t encourage you to use old versions of SQL, but I just like this little … I don’t know, it’s nice. It’s nice. I like nice stuff. It’s just touching. Mass produced, but still touching. So then you run set up. So you get your install media, you get it up there, you run set up executable, and then you stop. Seriously, you stop, throw it away. Don’t do an in place upgrade. If you were here hanging on the edge, just like oh this is great … I understand there are situations in life where you have no choice. You’re on a physical server. The management says, this is how we’re going to do our upgrade and we’re not buying a new server and we have no swing server, deal with it.

Yes, there are situations where an in place upgrade is your only choice. I’ve done them. I’ve helped people do them. They are so risky. I don’t like in place upgrades. As far as it’s up to you, put your foot down and say, we will never do an in place upgrade. Here’s the deal … if you’re on a SQL server 2008, when was it last built? If you happen to buy brand new SQL server hardware last year and you just moved your 10 year old SQL server 2008 to it, well yeah it’s new hardware. Maybe they don’t want to spend the money, and I understand that. But most of us are in virtual. Many of us are in the cloud. There’s other ways to do it.

We had one person ask about in place upgrade and we said, we don’t recommend it. Here’s all the reasons why, even though it’s simple. They went ahead and tried it and they said, nope don’t worry, we’re going to do it. We said okay, we’re support, you’re going to need it. 99%, maybe 95%, maybe 90, whatever, some percentage of the time. Not as good as 99.9, but some percentage of the time, they work flawlessly. When they go bad though, they go really bad. When they go bad in the middle of your upgrade, you have no rollback. Your rollback is to start mashing buttons to hope the old version works to try the uninstaller, and to do a bunch of stuff to your production SQL server that’s been your bread and butter database server for how many years. So I’m not of it. It’s pretty clear.

Again, they’re not that bad. They have to happen sometimes. If you want to see more, I’ll have a link for this. But if you just do a google search with DBA.stackexchange.com and you search for in place upgrade, you’ll find this answer I gave. It’s well uploaded. I have 91 votes on it. This is just one page. You’d have to press the page down button like four times. Anybody who’s in this webinar whose worked with me as an employer, as an employee, as a customer, you’ve sadly probably seen my long emails. I’m working on it, but some of my stack overflow answers can be long too.

My short answer is no, don’t do in place. If you have a question about in place upgrade, throw it in the comments and we can talk about it. We can talk about the live webinar, but we [inaudible 00:09:15] the Q and A in a couple weeks.

So the first thing we do is we’re going to install SQL server on the new. Hopefully you’ve actually done this in test already a couple times. We talked about this last week. You want to test, test your migration, test your compatibility mode, have your applications point to it, get a test version of your … Ideally, the ideal thing to do is for you, before you do your upgrade, to do an upgrade advisor, maybe to do the DEA tool, then maybe take SQL server database tools and run a build against it. Now that’s for your database. If it’s your vendor database, it might be as easy as talking to your vendor and ask them what do they support, and just cross your fingers, pray, meditate. Whatever you need to do, hope that your vendor is going to be compliant. Oh yeah, of course, we’ve already tested it and we certified it. You’re good it’s easy.

Sadly, a lot of our vendors just don’t play that game for us. Actually, I shouldn’t say a lot. Half do, half don’t. So it’s 50/50. But if it’s your own application, you want to actually run a build against it, against the target for the version you want to go and just see what breaks. But all of those tests alone are not as good as deploying SQL server X. So let’s say you’re on 2008 today and you want to go to 2017, upgrade to 2017 and test, deploy your database, get it all good there, and then point your test suite to it. Point your applications to it and do a full regression test and see what breaks. There is no better testing than a full regression test to see what breaks.

Once you’re done, build a new server. Build it right. So remember, you have an awesome opportunity here. You have a chance to take all the knowledge you’ve gained over the past 10 years and apply it to this new server. You can [inaudible 00:11:00], you can make sure power saving is enabled, you can finally have 64K drive allocation size when you format your drives. You can, if you’re on a VM, you can do the right data stores and para virtual scuzzy card connectors, and you can do all the best practices that you keep reading about in the SQL server community, but you say well I don’t want to touch it because it’s already here and we don’t want to break it. We want to wait for the next version. The next version is here.

You have no choice, by July, if you’re on SLQ server 2008 and you don’t want to spend a bunch of money on Microsoft and move to Azure and get their secret slug worth deal, or whatever his name is from Willy Wonka. But you have a chance now to build your SQL server to best practices. Run a health check on it, even though there’s nothing on it yet. Run a health check and see what settings you got wrong. Add maintenance. If you don’t have all the [inaudible 00:11:53] maintenance scripts or some trusted maintenance solution, install that in the new server and experiment with it. So when you migrate over there, you don’t have to say oh yeah … because here’s the deal, tomorrow is a big enemy of us. It’s an enemy of me. Tomorrow is the worst word in our career.

How many times have migration happened, and it finally gets done, and something goes wrong, a little bit minor but it’s not big. You get it solved at 1:00 in the morning and you say, alright, in the morning I’ll fix the backups. In the morning I’ll fix the index maintenance. In the morning I’ll fix this. Why do that? Set up your maintenance the right way today and have it scheduled automatically. It’ll be really quick right now when no user database is on it for the two weeks you’re waiting to move it migrate, but set it up head of time and then it’s just good to go. Set it up automatic is good. Setting up ahead of time is good.

Don’t recreate security woes. Does the whole entire world have SA access to your SQL server? If they do, fix it. Now, there’s two schools of thought there, I get that. Some people will say, well yeah, but we’re doing a migration. We want to break as few things as possible. I get that, sure. But if you can start over and have the best practices in mind, do that. Again, if we have a test first mentality, we’ll test this and we’ll see what breaks. Does your vendor really need SA access to your consolidated SQL server? Probably not. Will your compliance people like to know they do? Probably … Not even probably. They don’t, period, end of story. They don’t even like that you have SA access.

Again, you want to pick the right resource. So we talked last week about grabbing some perf data and doing the baseline analysis. If we do the baseline analysis and we know what we need, and we know what we’re using, we can actually build the new server to spec, and build it with the right amount of resources. Yeah, so Mark just mentioned a great comment. Mark says, “I have nothing but praise for Ola Hellengren, but make sure that’s his default are appropriate in your environment.” Yeah, absolutely.

We just had a conversation the other day with somebody on my team about this. Not somebody on my team, but somebody off my team. The defaults are bad. If you just take the default in the next maintenance job and run it, you’re doing the next maintenance. If you schedule it every night, like you just want to update your status, every night you’re doing an index rebuild and you’re not even updating your stats. So don’t just … True of any script and run it. You want to actually go to Ola.Hellengren.com, or whatever his website is. We just do a google search for Ola SQL and you’ll find him, but you can say Ola Hallengren, H-A double E-E-N-G-R-E-N and SQL, and you’ll find his blog, or his maintenance [inaudible 00:14:39] website really.

If you just click in any of the links for index optimize or backup, you’ll see pages of options. He even gives examples, so there’s really no excuse to run the defaults. Do not do it. Thanks Mark, that was a great point. Don’t forget to enable the maintenance clean up. So you have a new server. Keep your MSDB happy from the beginning. Do your MSDB cleanup history right away. Do your job agent history cleanup right away. Keep these things pruned and well oiled from the beginning. Recycle your air log on a regular basis. We’ll talk next week about some best practices in the after and kind of talk about some of the things that we find that health checks commonly, and getting your monitoring set up again. We’ll talk more about that next week.

Here’s a big thing. I’m a big fan of CNAMEs and aliases and not connecting to server names. If I have 10 applications on a consolidated SQL server, I might be as crazy as asking IT for 10 CNAMEs. Now you don’t have to go that far if you don’t want to, but I don’t like people connecting to the SQL by its actual host name. I want them to use a CNAME always. There are schools of thought. Is it silly to have CNAME if it’s going to an AG listener, or if you have a cluster incidence, because that’s kind of a C name going to a CNAME. Yeah, maybe, but I like them still. This is because I’m thinking my future upgrade, right?

SQL server is now coming out every couple years. Their servicing model is different. Staring 2019, there’ll be no more service packs. We’ll just have CUs and slip stream CU updates. So SQL updates come fast and furious now. You don’t have to take the train every time it comes by your stop, but you shouldn’t be four trains behind, like many people are today, or five. So, if we’re going to start upgrading more frequently, we want to make our upgrades easier. If we take the pain today and make a CNAME and migrate to it, that does two things for us. We’ll talk in a second when I talk about operation honey pot and how that can help us, but it also makes future migrations easier.

So you’re making today’s self mad. This is stupid. But in two years, your future self will say thank you. I could sit here for hours and talk about all the ways that old Mike screwed over today’s Mike and I hate it. Right here at the farm, we’re running out of our fire wood. We probably burned five cords this year. It’s been so cold a couple times and we’re not using oil. I cut my wood way too late, so it’s way too green and I’m probably going to buy a cord of wood this weekend and it’s going to be way too expensive, but I’m stubborn and I don’t want to burn oil.

I’m mad at the Mike who was sitting on the couch last spring or two springs ago instead of cutting down woods here at the farm. Take the time today to make your future self like you. It will be amazing. So make the CNAME. When we talk about operation honey pot in just a few slides, it will give you another tool to find those non-compliant, frustrating developers and vendors. So we’re not going to read all the slides here. You can get the slides on the blog post when I share it next Tuesday, but it starts with a plan, right?

So you’ve got to have an upgrade checklist. I love checklists. I don’t use checklists nearly as much as I should. I’m on the fire department. We have some checklists there for truck checks, and we have checklists in the ambulance for certain cardiac emergencies. Checklists are a really important tool. Like I said last week, if they’re good enough for surgeons, if they’re good enough for pilots, even big restaurants, chefs use them. If they’re good enough for all these other industries, I think they’re okay for us to use too.

So all these are the kind of things that go on a checklist. Again, I’m not going to talk about these, but think about the things that we want to do before. So there’s kind of three components to a checklist, just like the three components in my..

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

This article is part of our SQL Server Upgrades Made Easy series:

  • Why Upgrade?
  • Before You Upgrade
  • During Your Upgrade
  • After Your Upgrade
  • Recap Chat
Upgrading SQL Server – Week 2 “Before you Upgrade SQL Server”

You can watch week one here on the blog or directly on youtube. Please consider subscribing to the channel as we’ll make it a rule that all webinar content from the free webinar series will be shared on youtube and sometimes here on the blog also.

This is the second video from the series from the webinar last week. Here we chat about the things to consider BEFORE you start your SQL Server upgrades.  We discuss questions you should be asking yourself before any SQL upgrade project starts. We talk about rollback plans, CNAMEs, checking health before, etc.

Before you Upgrade Your SQL Server - SQL Server Upgrade Webinar Episode 2 - YouTube

Sign up for the rest of the series and to get the full cuts from Webinar Ninja with any Q&A each week. Subscribe to the youtube channel to be alerted of the live chat hangouts and to see the upcoming talks in our continuing webinar series. In March, I’ll start with a multi-part series on High Availability/Disaster Recovery options for SQL Server. If you have questions, you may not need consulting. Just check out the series, attend live and ask your questions or come to the Q&As. I’ll try and answer most questions left in the posts also. Yes, I own a consultancy, and if you need in-depth help, I think we can help – and I trust you can figure out how to reach us. But the point of this content and the webinar series is to help encourage you and show you how can you do more than you realize! See you on Thursday 2/21 for part 3 on “Doing your SQL Server Upgrade”!

Links Discussed Transcript

(I’ve used a transcript service for this and I reviewed most of it. It’s a work in progress and there may still be some errors!)

Welcome back. Last week we sort of talked about why should you upgrade and the real answer is “why not upgrade?” I mean SQL Server has changed so much since the SQL Server 2008 that you’re running or the SQL Server 2012 or 2014 even and hopefully not the SQL 2005 and 2000, hopefully you’re off that. So I think we all understand the why, it’s really why not. But the world has definitely changed as we talked about last week.

So what I want to talk about this week is a little bit more about “the before.” What are some things to consider? And this is going to be not demos, it’s really going to be chats so if there’s questions ask them live in the chat, if it’s germane to this topic we’ll kind of put me off topic and we’ll talk about it. We’ll keep this 30-35 minutes. I have a little more slides than maybe that time so if we need to we might go longer or I might just cut myself off and extend more next week.

So anyway we’ll go in here. So again this is the series. Last week was big picture, kind of a “why why why” and what’s changed. And I gave you some homework to basically analyze the versions that you think you want to be at, analyze why you wanna upgrade, maybe start conversations with your vendors because the vendors are the pains here and they’re the ones preventing you from upgrading in a lot of cases, talk to the company and really see what are you going to do. I mean especially if you’re on SQL Server 2008, 10, 11, 12 year old database technology that’s gonna be end of life, with all those caveats we talked about last week where Microsoft is making some special deals. If you just give them more money they’ll still give you some patches or if you go to Azure they’ll give you the patches for free. And I talked to you about my opinion of that. So it’s time to upgrade. And the homework was really to start getting yourself ready and getting your organization ready.

Today we’re gonna talk about what to do before, how do we plan our upgrade project right. We’re about to do a massive change to our SQL server. It’s not easy and it’s not small taking a SQL Server that’s running really well here for 10, 11, 12, 14, even 7 or 4 years, and move it over here to a totally different SQL Server version, to a completely different infrastructure potentially, upgrading to a version that you’ve never been on. The world’s changed. You’ve been so busy you’ve not learned new stuff. So what are some things we can do to make this successful and that’s what we’re talking about.

Again I’m giving you something that could probably be a 30 or 30 hour course and a lot of consulting. I’m trying to give it to you in a handful of 30 minute webinars. So we’re not going to cover everything but we’ll get to a lot. I already told you who I am. I’m a guy who’s a SQL MVP, I’ve been doing SQL for 20 years, I’ve been consulting for 10. I run a company called Straight Path Solutions. And more importantly that’s my blog. I’ve been blogging there before I was consulting there and so the company name just formed after my blog name. I’m a user group leader and I have a lot of scars from SQL Server upgrades. In fact I used to have a full head of hair before I started upgrading SQL Servers.

So this is a good quote from a poem. It’s, “The best laid schemes of mice and men” and then it’s said in like Old English. But it’s “The best laid schemes of mice and men often go awry.” So we need to be prepared. Louis Pasteur says “Fortune favors a prepared mind.” When we sit down to begin any upgrade project we want to make sure that we can answer a bunch of questions.

And what are the questions? Well, half of SQL Server or more is just basic human nature stuff. So can you answer the who, what, when, where, why and how questions. And if you can answer those questions you’re halfway to having a good upgrade. A lot of times when we go in as consultants and help somebody do an upgrade it’s not really going in and doing a bunch of really insane and in-depth analysis. Yes, we use our SQL Server knowledge but a lot of times it’s really half sort of what I call “business therapy”. We’re just helping them ask these questions. We’re just coaching them along on saying “What are you trying to do, where are you trying to go, why? Who’s in charge? What team do you need to have in place? What are you gonna go to? So what version of SQL are you going to go to? What’s the problem you’re trying to solve.” That’s also the why question and I have a capital Y in the W slide, I’ll have to fix that later. “When do you want to go there and when are you gonna test it and how much time can you be down for? How can you do it with minimizing that? What are the risks? How are you going to do it?”

Ron says, “Why are you still on SQL Server 2000?” Yes, we’ll talk about that in a couple slides. But absolutely, “How do you roll back.” And next week we’ll talk about it.

I absolutely hate, and I’ll use that word, I hate in place upgrades. In place upgrades when they work are OK and it’s like “Phew, it worked.” But even when they work we’re still on the hardware we were on. We’re on the same OS we were on. We’re still living on a bed of the stuff that we did five years ago or 10 years ago, or in Ron’s case, 19 years ago. We’re still, we’re living on that bed and if we learned anything and hopefully you’re learning something every year … I mean I wish I could go back to a company I used to work at full time and say “Oh I wouldn’t have built like this, I wouldn’t have made the config like this, I wouldn’t have done the operating system like this. I would have done 64K drive allocations instead of the default.” Because 15 years ago I didn’t know about that. I would have built with the CPU configured this way. I would have had less CPUs with better performance and I would have potentially had and standard editions of Enterprise. All these mistakes your current environment is built upon. And they’re not mistakes that are bad per se, it’s not that you’re a bad person because of them, it’s just you didn’t know.

So you know much more now. You’re going to webinars with some guy from New Hampshire. You’re in the SQL Server community. You’re taking classes. You’re reading books. You know more this year than you knew last year. As a consultant I do that all time. I’ll go back to clients, “Hey listen, remember that thing I told you do three years ago? That was good three years ago but we can do it better now and here’s how and here’s why I learned it.” So even if an in place upgrade succeeds I don’t like it for that reason. But when in-place upgrades go bad and “best laid schemes of mice and men often go awry.” Well the best laid plans of in place upgrades can go awry and when they go awry they go awry bad and the rollback, that’s not a fun weekend. It’s like an “I hate my job. I’m updating my resume” kind of weekend. So we’ll talk next week about why I prefer migration style upgrades. But the biggest reason is rollback is just as easy as changing your CNAMES and we’ll talk about CNAMES in a few slides actually.

But it’s as easy as changing your CNAMES, pointing back to the old server and saying ‘I don’t know what happened. We had a problem.” It’s Sunday night at 2:00 in the morning or Monday morning at 2:00, I don’t wanna deal with it anymore. We’re rolling back. We’ve made the decision. Because when you start thinking about your who, what, when, where, why and how you’re going to ask yourself questions like how do we know we’re successful, at what point do we say we’re we’re putting the stick in the sand and we’re saying, “This is the line.” Where’s your red line and how do when you’re at it and who’s watching it? Because as technologists, I have a blog post out there, maybe Mike and my team will find it, but I basically talk about don’t forget to fly the plane. And I talk a lot about aviation disasters as well. As technologists and engineering type people we love to get into the nitty gritty and we might get so lost trying to fix a problem that the sun starts coming up on Monday morning and “Oh no, we’re up the creek.” So you have to have somebody watching the time and being willing to say “No, we’re rolling back.”

And with a migration upgrade rolling back as you change your CNAME over you say “OK, we didn’t hit the upgrade this weekend.” Instead of saying “Oh no we’re in trouble” and then feeling bad and sort of getting dejected, you say “We failed. Great. We learned what didn’t work. Now let’s go forward and learn from those mistakes and make it even better try next week.” Yes we had a downtime window. There’s a little opportunity costs or some sunk costs there that you might get worried about, “Oh no they’re going to hate me.” But your users I guarantee you will hate you a lot more if you kept going and going and going and you lost data.

“Do you have any specific steps, plans, tasks in Excel you’ve prepared?” Yes actually. I’m working on a document for that and we’ll probably have that checklist next week. So that’s one of the downloads that we have a slide already talking about. But again I can’t give you all of the steps but I’ll give you some of the steps that I like to use. So that’ll be in a download next week with a session next week on the doing but definitely have a checklist. Checklists are really important and we’ll get to that in that slide. So I’m gonna keep going because if I stay on one slide I’ll get lost but ask questions and let me interact. Again, the whole purpose of why I’m doing this and why I want to continue doing webinars for as long as you keep coming to them is to interact, to ask questions, to challenge me because I’m going to learn from you. So let’s just keep the dialogue going if you have questions.

So this is another good quote. “To remember where you come from is part of where you’re going.” So that quote should scare you. I bet you Ron, like Joseph from last week and judging from you talking about SQL 2000. I bet that quote scares you. You need to know where you are. You need to know what’s good in your current environment, what’s bad in your current environment. So even before you start with a migration plan, even before you pick your hardware or pick your cloud provider or pick your virtualization layer and spec out your servers, you need to know what’s good and what’s bad in your current environment.

So again, I’m not trying to talk about my company to try and sell my company. I’m talking about my company because it’s an example of how I do this all the time for customers. When we do an upgrade and you can do this yourself with tools like Glen Berry’s Diagnostic Scripts, Brent’s Split Script, you don’t have to call us. We don’t need to do your health check. You can do your own health check. You’re here. You figured out how to come here. Send me an e-mail. And in fact I’ll give you a download in an e-mail after this that links to our self-assessment PDF that you can just take on your own and not have to give me your email address for. But the point though is, do some assessments, look at your environment and understand what’s good and what’s bad and then look at your performance.

I have a blog post that talks about using the PAL tool, the performance analysis for logs tool, that’s a free Perfmon based tool that can do some analysis and you can use as a baseline. If you have Century 1 or Red Gate SQL Monitor or any of those third party tools that do monitoring. The point is to start doing some monitoring. Look and see what your resource utilization is. But don’t just look at your resource utilization, also look at your queries, look at your settings are you using more resources than you need to because one of your feet is on the gas and one of your feet is on the brake. So are you hurting yourself and is that making you use more resources than you need to.

So do this health assessment and every time we do an upgrade or migration I do a health assessment first. And sometimes the customer will say, “Well why would you do a health check for us? We’re getting rid of the server.” And I do the health check first because I want to see what you have and I want to see what’s different and what’s unique about your current environment. Because maybe you forgot that you have a default collation of something other than the normal default collation in your instance and you have four databases that have it this way and two that have it this way and you forgot that “Oh yeah I have this linked server that goes to this DB2-AS400 environment or something. And by doing your health assessment you get to analyze what you have and make sure that you’re thinking about all these things. So the health check really gets you in the mindset of sort of doing a, for lack of a better word, a tabletop exercise of thinking about when I do my upgrade what’s going to go wrong, what’s going to go right, where am I hurt today.

So a couple of things you can do is do a health check, do some performance analysis and don’t forget to look at the licenses your using. If you’re on 2000 or like Ron said if you’re on 2008 and you bought your licenses earlier with software assurance the licensing scheme has changed.

So do you need to be on Enterprise? We talked about that last week, I won’t rehash that. But SQL 2016, actually I will rehash, I’m about to. SQL 2016 SP1 gave us a lot of Enterprise features in standard. SQL Server 2014 and higher we can use 128 gigs RAM in standard. I’m in the middle of teaching an HADR class and some of my students are here from that class this week. You can do a failover cluster and then you can do a transactional replication for reading and you can have H.A. and you can have a read only replica through the replication without having to buy Enterprise Edition to use availability groups. You get failover cluster for H.A., you get replication for reading and you can do log shipping for D.R. and log shipping is perfectly respectable. When we do our webinar series on HADR I’ll talk about that. Log shipping deserves more respect. So do you need to be Enterprise Edition.

And don’t forget to look at your licenses and look at what kind of CPUs you can do and don’t just buy. “I don’t know what we’re using and I hope we don’t have a bunch of performance problems so let’s go with the biggest and beefiest box and get 482 cores just in case we need them.” That’s going to hurt you. So we want to try and be a little more spartan.

Now if we’re building on physicals we might want to get it closer to being right because on a physical we can’t just go and change CPUs if we realize that we got it wrong. On a virtual that has proper allocation and has enough headroom of more CPUs we can get it a little bit wrong and add CPUs. So I would start smaller CPU core wise and do some performance testing before you upgrade and watch and see what you need and we can add CPUs.

And then the cloud. If you’re building a VM in Amazon or in Azure you can get it wrong all day long and just reconfigure your server and have a brief blip and change your server from a small box to a gigantic box with minimal effort and minimal downtime. So we can get it a little bit wrong but we want to do this performance analysis to get it kind of right.

So another good quote that I like is, “If you don’t know where you’re going you’ll end up someplace else.” Isn’t that true? I mean and I’m preaching to the choir but how often do our companies do this to us? How often do our managers and our project managers and our vendors give us a marching order, and hopefully nobody from my team says, “All the time.” But give us a marching order and say, “We’re going this way, we don’t really know where we’re going but we’re gonna go this way.” And there’s no plan. There’s no goal. There’s no, “Here’s where we are now, here’s where we’re going. And this is the road to get there. This road can’t get there. This road can’t get there.”

Another saying out there is, “If you don’t know where you’re going you’ll end up there.” So again from the homework last week did you pick a SQL version? Again, I prefer SQL Server 2017 I think for most new environments that are calling me. 2019 is kind of around the corner but just based on my experience I think 2017 is a good place for most people to go. Even 2016 is an improvement over your SQL 2000 and 2008. If your vendors are a little slower and they’re getting nervous about 2017 you have more time before end of life for those two.

Are you going to new hardware? Hopefully you’re not doing an in place upgrade. Is it time to move to a virtual environment? Is it time to look at the cloud? And we’ll do a little sidebar chat at the end here if there’s time for the cloud to talk about that. But ask these questions today. Don’t ask these questions one week before go live, or even one month before go. Ask these questions now and figure out what are we about to do and how are we going to get there and why. And again go back to the questions I asked earlier; who, what, when, why, where and how. And all those questions, as you start asking them over and over again, and you could even ask underneath so like why, why, why, why? Like ask the why and then ask the why about the why and keep asking whys until you get to the discrete tasks and discreet questions and worries you should have and these become your checklist. And again we’ll give you a simple checklist next week. But that checklist is not going to be end all be all. It’s not

going to be end all be all it’s, not going to be your checklist; it’s my checklist and it’s generic, and you have to fill in the details and the gaps, because your environment’s different, and that’s why you have to spend this time not doing work, not troubleshooting problems, but asking these questions and thinking about your upgrade.

I like this quote too, “The most exciting phrase to hear in science, the one that heralds discoveries is not eureka, but now that’s funny.” That’s from Isaac Asimov. We got to test; the point here is we have to test. We shouldn’t … It’s exciting but not really a good exciting, necessarily, when it comes to a SQL server upgrade. That’s a cool discovery, like when you’re playing with moldy bread and you discover antibiotics, it’s cool when you discover how radiation can be used for medicinal purposes and discovery purposes; it’s not a great thing to say at T-1 second, before go live in the middle of an upgrade like “Oo, I forgot about compatibility level. Oops, I forgot about the encryption. Oops, I forgot what the link servers. Oh, no, this version of SQL in this compatibility mode, do not work. I should have known that ahead of time.” You don’t ever want to say “Now that’s funny.” During your upgrade.

So we need to test, we test two things. And really not just test two things, we also need to talk to our vendors and get our vendors to bless us; that’s hard. We talked last week about it, and if you want to vent about vendors here, you can, just remember this..

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

Upgrading SQL Server – Week 1 “Why, Why, Why!”

This is the first video from the series from two weeks ago. This is an introduction to the SQL Server upgrade series. We also start to get into a few of the questions to consider before upgrading – primarily the WHY of the upgrade. If you are on SQL Server 2008 or 2012? It’s definitely time to consider an upgrade. If for no other reason than all of the new features. Not to mention, SQL Server 2008 is going end of life. This series is really designed to help you ask the right questions, get into the right mindset, plan for your environment and give you some tips and tricks from all of the SQL Server ugprades I’ve been a part of as a consultant and full-time DBA. Enjoy! And feel free to post questions!

Why Upgrade SQL Server? SQL Server Upgrade Webinar Episode 1 - YouTube

Sign up for the rest of the series and to get the full cuts from Webinar Ninja with any Q&A each week. Subscribe to the youtube channel to be alerted of the live chat hangouts and to see the upcoming talks in our continuing webinar series. In March, I’ll start with a multi-part series on High Availability/Disaster Recovery options for SQL Server. If you have questions, you may not need consulting. Just check out the series, attend live and ask your questions or come to the Q&As. I’ll try and answer most questions left in the posts also. Yes, I own a consultancy, and if you need in-depth help, I think we can help – and I trust you can figure out how to reach us. But the point of this content and the webinar series is to help encourage you and show you how can you do more than you realize!

Transcript

(I’ve used a transcript service for this and I reviewed most of it. It’s a work in progress and there may still be some errors!)

This is really what the series is about. You’re on day one, and you can see the agenda here. The main goal is not … I’m not going to upgrade your SQL Server here in 20 to 35-minute webinars. I’m not going to share a lot of code because I think a lot of code just adds complexity that you don’t need. My main goal here is really to just help you think about SQL Server upgrades.

I’ll tell you who I am in a second, a little bit about my bio, but the short story is I do a lot of SQL Server upgrades. I’ve been doing SQL Server 20 years, and I’ve been consulting for the past 7 or 8. So I have the benefit of, as a consultant, getting called in a lot for upgrades. So while you all may do one upgrade every 7 years or 5 years or 3 years if you’re lucky, or unlucky depending on how you look at it, although hopefully after this series you’ll say lucky, I do upgrades all the time! So I get to see the good, the bad and sadly I get to see the ugly. Sometimes, as a consultant, I get called in and I see the really ugly and I want to help you avoid that. I want to help make you think … This is more of a watching my hands move, watching me talk, looking at some slides and really help you think of the things that you should think about.

It’s one of those things that you might do a normal regular old DBA, that my first 10 years of working with SQL before I was a consultant, I think I did two SQL Server upgrades. A normal career span you might, if you’re at one company, just do 3 or 4 upgrades, maybe 5 if you have a lot of SQL Servers. I want to help you think about other ways to do it, how to think about downtime, how to think about versions. All the questions that are in your head I want to help answer in this webinar series.

So today’s big picture. I want to chat, introduce you gently to the webinar series and we’ll talk about why, why, why. Why upgrade? Why haven’t you? And why should you? What’s changed? What’s new since your last version of SQL? And next week we’ll talk about some more before items. We’ll say, “What should you do before you upgrade?” Before you even start with the new server, build a VM, go to the cloud, whatever you’re doing, what do you think about before? How do you plan? How do you plan for the downtime? How do you communicate? How do you test? All those questions.

Then the next week we’ll talk about doing your SQL Server Upgrade. So while you’re doing your upgrade, how do you do it and manage success? How can you do it and minimize downtime? How can you do it and come out the best way possible and not be scarred for life?

Then we’ll finally talk about after the upgrade. So once you’re done with your SQL Server upgrade, it’s a tremendous time for you to start doing best practices. It’s a tremendous time for you to start saying, “We’re starting fresh.” It’s like New Year’s Eve or New Year’s Day, you’re starting fresh for the year, you start a new job. When you have a new anything you get a chance to do things a bit different than you did before. So I want to help you in that last webinar talk about how do you start off on this new version of SQL on the right and best foot? Frankly that last webinar will be good whether you actually do the upgrade homework and upgrade or not. It’s just what’s some good best practices for your SQL Server environment?

Then, finally, the first Thursday in March, on the 7 of March, we’re going to do the whole StraightPath team will be on a Google Hangout with you, so we won’t be in Webinar Ninja. We’ll be on a Google livestream Hangout and we’ll just be there to ask questions, answer questions or maybe the team can make fun of me for things I say in this webinar series.

So that’s a high level overview of what we’re going to go through, and thank you so much for being here. I really appreciate it.

Who am I? Not that it really matters. I would put a picture there but you can see me right here. I’m a father, a husband, I have a small family here, 4 kids and I’m a farmer too. We have cows and pigs and chickens here up in New Hampshire. Just had a new calf born. So if you look at my Twitter account it’s @mike_walsh, you’ll actually see a picture and video of the new calf that was born on the farm. It’s fun. Just a little Mike trivia in case you ever get asked.

And I’m a founder and I’m a friend. I’m a friend to a lot of people and that’s what this is about. This is not about StraightPath, this is really about me, Mike Walsh, SQL Server MVP, just trying to give you some tips and tricks and things that we do at StraightPath to help customers. I’m trying to give you some of our secrets and some of the things that we do to make upgrades go successful.

StraightPath, if you want to see our blog, it’s straightpathsql.com. We don’t have any downloads for this week but next week and the subsequent weeks to that we’ll actually have a couple pieces of downloadable content. It can be some checklists, some how to guides and some lists of questions. Just some things for you to take and walk away with and have a tangible asset to remember the things we discussed.

I run a user group for SQL Server and if you’re here and you’ve not been to a SQL Server user group, I suggest you just go to the PASS website, PASS.org, and look for your local user group. There are so many great sessions out there, so many people who do what I’m doing right now, who just give up on their time and say, “I’m going to help you.” And they don’t want anything, none of us want things. Sure, we do consulting and if you need help I’m willing to help you, but I don’t expect anything in return for this. I just want to help you and how that’s the whole SQL Server community is. That’s how I got my start 10 years ago, people gave back to me early on through going to PASS and reading SQL Server magazine back when it was in print and user groups and answers to questions on the news groups, if you’re old enough to remember news groups. It’s inspired me to give back and that’s what the whole SQL community is about.

Upgrade scars, I definitely have a few. And I think that’s good because the more you do things wrong, you make mistakes, the more I think you learn and get better. I want to transpose my knowledge of my mistakes onto you so you don’t make the same mistakes.

Let’s talk about this week’s webinar. Again, this is the high level introduction overview webinar, but I want to answer a couple why’s. Why should you upgrade? I’m not going to play with polls, because polls I have to scroll down and all that, but if you want to type, you don’t have to, in the chat window, I’m just curious if you’re on SQL Server 2008, just type something in the chat window and I’ll see it out of the corner of my eye. A lot of people are still on SQL Server 2008. Yup. I know you are, Ron. Hey Ron.

A lot of people are still on SQL Server 2008. In some ways that’s a testament to how great the product is, but it’s also a testament to how busy we are, how maybe scared we are of upgrades, how frustrating our vendors can be. It’s a testament to a lot of different things, both good and bad. But SQL Server 2008 is end of lifing in July, I think we all know that. So it’s already end of mainstream support and now it’s going to be end of extended support. Yeah, sometimes management refuses to upgrade, Dale, and hopefully we can help you with that.

Basically, here’s the big thing, if you’re in an environment that’s dealing with compliance and security and audits and you’ve not upgraded, after July your auditors will start slapping your hands. As I’ve been consulting for the past 7 or 8 years, I’ve noticed the compliance auditors are getting scarier and scarier and I’m finding more and more companies saying, “No, whatever the auditor says we’ll do.” I don’t know what industry you’re in, Dale, but that can be a carrot and a stick to get them to agree.

But a lot of management refuses to upgrade for license price reasons, we’ll talk about that in a couple slides, we’ll talk about that more next week. A lot of times they refuse to upgrade because there’s a fear of what’s going to break. It’s this monolithic thing, it’s this big, heavy thing, your SQL Server, with a ton of inertia, and you’re trying to migrate it from here to here on a new version with all these things changing and live users and all this and management goes, “I don’t want that on my resume. I don’t want a failure.” I totally get that too.

So that’s what happens. A lot of people are on Windows 2008 or Windows 2012 … Yeah, 21 terabytes, that is a scary size to upgrade but it doesn’t have to be. We’ll talk about that and how to manage it.

I will tell you we’ll talk a little bit next week and during the doing that I am radically opposed to in place upgrades. So if your management says you can do a migration but you can’t do an in place, I agree with them. In place upgrades scare the living daylights out of me. They go really well 85% to 95% of the time. It’s that 5% to 15% that gives me night sweats.

If you’re on SQL Server 2008, after July you have no more support. Caveat time, caveat, Microsoft says, “We can get a little greedy.” So they’re basically saying that if you want to stay on SQL Server 2008 and you want to get important security updates, we’re going to back down a little bit. You can do that, you just have to buy software assurance and you have to pay for it. You have to pay basically … It’s basically like buying a new SQL Server 2008 license, or really probably a 2012 license or the current licensing scheme, every year. If you pay for that every year we might give you some upgrades.

And now Microsoft is saying, “But wait, act now and you can get a special deal.” It’s like those infomercials of Ron Popeil. They’re saying, “If you just go to Azure, we will give you that thing, the free updates or the included updates, for free. Just come to Azure and we’ll give you it for free.” Don’t take that offer. I forget the guy’s name in “Willy Wonka and the Chocolate Factory,” but that’s one of those bad deals. Azure is great, I love Azure, we have a lot of clients there, NAWS, but it’s a bad deal because it’s SQL Server 2008 and the world has changed. We’ll talk about that in a couple slides.

This is also a chance for you to upgrade your hardware and your VMs. We’re talking to a client who’s doing upgrades right now and they’re on 2010 era CPUs. We have more cores than they need. I bet you we go to newer CPU, we can downgrade their cores, save them on licensing.

If you’re on SQL Server 2012 or SQL Server 2014 and you just went, “Phew, Mike’s not talking to me.” No, I’m talking to you too. SQL Server 2012 and 2014 were great versions of SQL Server. 2012 gave us a lot of new features, availability groups are a big one, but Microsoft V1 is always, hey, it’s V1. You never know. V2 of all Microsoft products tend to be better. And if there’s no V2 of the product you know it was not really a good product to begin with by and large.

So SQl Server 2016 and 2017 are really more powerhouse versions of you and there’s a lot more features out there. We’ll talk about that a lot in the rest of this webinar series. But if you’re on 2012, 2014 or 2008, it’s really time to upgrade, which is pretty much why you’re here.

So why haven’t you? Dale says because management refuses to upgrade. That’s some place in there. But these are just some of the reasons. I’m not going to read the slides to you. These are the feelings that people get in your head, like, “OH No! Where do I start? What version do I go to? What’s going to happen here? Do I stay clustered? Is it going to break? What about my connection-

Do I stay clustered? Is it going to break? What about my connections? I don’t want to have to go in and change IP addresses. I don’t want to have to go into all my ODBC connections on these thick client apps I have all over my entire organization. I’m just going to stay where I am and clench my fist, cross my chest and say, “Forget it.” A lot of people get into that. These are some of the fears and I’m sure you have your own. I’m sure if you wanted to, feel free to type out your fears that I’m missing, but I think the last one is the big one, the big catch all. What questions am I forgetting to ask? What are the unknown unknowns? What will break? Or for a 21 terabyte database, how do we migrate it and do it quickly without having to have two weeks of downtime?

Yeah, Ron says, “That’s a great point. Small vendor company doesn’t have proper SQL knowledge and they can’t even explain if they support it.” Yeah, call up your [inaudible 00:13:55] log shipping is what we do, exactly how we do it. So it’s less scary and we’ll talk about that in a couple weeks in the doing upgrade webinar. But absolutely, Ron, vendors are a big part of the problem.

Some of them either say, “We haven’t certified that yet,” especially in healthcare.

Some of them say, “I don’t know, try it.” When you get a small vendor that says, “I don’t know, try it,” if it’s not going to violate your support contract try it. And we’ll talk next week and in the doing webinar as well about some things you can do. You can look at some extended events if you’re on 2012, you can actually do a database build. You can do some of your vendors’ work. You shouldn’t have to but you can do some of your vendors’ work to help make sure you’re going to have success.

We’re never just upgrading and saying, “Poof, it’s done,” we should always be testing. Always, always, always test. I think as long as we have that testing methodology and that testing mindset we should be good. These are the kind of questions that I really hope to answer in this series.

Again, these are going to be quick 20 to 35-minute webinars. I can’t answer all the questions but I just want to get your mind going about these things, and then you have my email address. I truly mean it, send me an email and say, “Hey Mike, I attended your webinar, I don’t need consulting help, I just want a quick question.” I’m going to answer your question because it’s fun for me.

So why will this series help us? I think I’m going to answer those questions that we just asked. I’m going to try and get those questions answered. Some of the questions, like why is my vendor a pain, I can’t answer for you. I can answer, if you call me up sometime I’ll tell you how I really feel, but I can’t make your vendor better. But I can help equip you to work around them. Get a new vendor. That’s easy for you and I to say, the DBA people.

Actually I have a great … I’m going to write this down and I’ll send this out in a recap. I have a blog post that I wrote 9 years ago, so you’ll have to forgive all the grammar and look. Give me a second. Blog post. I’m just writing this note, otherwise I’ll forget to do it. I have those blog post that’s called, “DBA Questions to Ask a Vendor.” In fact, somebody on my team might even be looking for it already. If you just google search for DBA questions to ask a vendor you’ll find that blog post. And if you do find it, go ahead and share it, Mike or Sheila.

The main point of my blog post, one of my points, was that you should, as a DB, ask a bunch of questions. They’re old antiquated questions, availability groups didn’t even exist yet, but one of the main points I had was the earlier you as a DBA can get involved in the vendor conversation you have a chance. Because every software vendor … Thanks, Ron. That’s old. Actually the post was actually from 2009, but I think I reedited the link for some SEO reasons. But I didn’t edit the content too much.

If you can get in the conversation early and have a sway on the vendor, it’s magic. I find vendors before the paper is signed are willing to do things. Vendors after the paper is signed are not. So sometimes you’re stuck with the vendors you have.

And we can help the vendors be better vendors. That’s something we do a lot. I’ve done performance tuning. I have a lot of vendors as clients now because we’ve come in for a customer of ours and we’ve had a conversation with the vendor to say, “Hey, listen, your product stinks for these three reasons. Here’s how we can make it better.”

Some vendors get mad and say, “Don’t help me. What are you doing? Leave me alone. You don’t know anything.”

Most vendors say, “Cool, wow, this is great. I didn’t know this. Thanks for helping.” So you can do that with your vendors potentially. Every vendor is different as you well know.

I think it will help because I’ve been there, I’ve been burned by bad upgrades, I’ve been called in … Our company has been called in when people have tried an upgrade and said, “Don’t know what happened. We just broke. Can you help us last minute?”

I want you to leave with a roadmap, and again we’ll have some downloads next week and the week after and the week after. I want you to create a roadmap in your head and on paper and I think that’s the main thing I want to do here. I want to help demystify this and help you. Again, here’s my email, write it down. If you get stuck just ask me a question. My rule is if you attended a session of mine at a SQL event if you’ve come to one of my free webinars, if you’re a former client, if you have a quick question and it doesn’t hit the 15-minute window I’m not going to worry about paperwork and billing. If you call me 4 times in a row with 14-and-a-half minute questions, by the fourth or fifth call I’ll get suspicious and send you an invoice, but I’d rather just help you.

So, anyway, enough blabbering, let’s get … So, the world has changed. And, by the way, as we’re going, if you have questions go ahead and throw them out here. We’re going to finish up in about 15 minutes at the most. If you have questions I’ll take some time with the questions.

But the world’s changed. I’m assuming you all know this, looking at who’s here and looking at some of the profiles. SQL Server 2016 SP1 gave us enterprise features and SQL Server Standard. If you’re a software vendor if you’re a customer who says, “I could really use compression but, man, I don’t want to spend the money on enterprise,” all the sudden standard edition becomes pretty appealing to you.

We have a couple customers right now that are enterprise edition who we’re going to downgrade to standard edition as part of their upgrade. That right there is a huge sell to management. That was a 21 terabyte database deal, I don’t know what the database is doing, but you’re probably going to stay on enterprise edition. Just a guess.

But SQL Server 2014 gave us 128 gigs of RAM in SQL Server Standard. Just those two things alone are big but we have availability groups, we have basic availability groups now in 2016, which means we can do availability groups in SQL Server standard. There’s a few caveats, you have to have one DB per AG, but if you have 10 databases or 5 databases you could have 5 or 10 AGs. It’s more to manage but it’s a viable option.

The cloud really is a viable option. We have a lot of clients on AWS and Azure, both IaaS and Paas, we’ll talk about what that means next week. And the HA/DR options out there are legion. You can just do availability groups, fill over clusters, we can do … I have a lot of clients that are on SQL Server standards with fill over clusters without shared SANs. They’re using third party tools like SIOS DataKeeper or in Windows 2016 we can just use the storage replica technology to basically make local discs on this computer and local discs on this computer appear to be in the SAN. And it’s supported and it works well and it’s performant, that’s not a word but..

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

Last week, I wrapped up the delivery of a 4-part webinar series giving tips about SQL Server upgrades. I talked about how critical a SQL Server Upgrade Checklist is to the success of your SQL Server upgrade projects. I promised I’d come up with a working basic example giving you ideas to make your own checklist with. Here it is! This is a free download, no e-mail address required. I hope you use this to help build your SQL Server upgrade checklists and have successful upgrades!

Why a SQL Server Upgrade Checklist?

Frankly, why not?! Pilots use checklists. Surgeons use checklists. The military uses them. Checklists help make sure no routine steps are missed, and they can make life a lot easier when you end up in an abnormal situation! Sure, at Straight Path, we do SQL Server upgrades a lot – we are consultants, so we work with a lot of clients. But for the average DBA who stays with a company for a long time, you can’t go wrong making your own SQL Server upgrade checklist!  

When an upgrade goes bad? It can get ugly quick. Your checklist will keep your upgrade running smooth. We remind you of a few things like having the right team in place, having a rollback plan, setting up the right order of operations. Combine the SQL Server Upgrade Checklist with our SQL Server Build Guidelines (JON LINK!!!)  and you’ll be off to a good start.

So… Start with our SQL Server Upgrade Checklist if You Want!

I told you that you need to have one. I hope you agree. Rather than pontificate on the reasons why, I’ll give you a free “start” – you can download Straight Path’s SQL Server Upgrade Checklist. You’ll note the disclaimer in the download and below. This isn’t the end-all-be-all for your SQL Server upgrade steps. We can’t write a SQL Server upgrade checklist that covers every single scenario. This is really meant to be a start for you to start your own checklist. Just use it. You don’t need to mention us, steal what works, test it, throw away what doesn’t. Leave a comment with things you’d add and we’ll periodically update this document and maybe keep a running (UPDATED: ) log at the bottom. And maybe make a new post when we do update it so if you subscribe to the blog, you’ll see it.

Our goal here is to help you out where you are. Sure. We do consult for SQL Server upgrade projects. You can always set up a free meeting to chat with me if you think you need a little help. But seriously, evaluate your needs, do your checklist yourself and see if you even need the help first! You may not!

An Excerpt from The Checklist

From the intro. There are very few paragraphs in this. It’s mostly checkmark bullets!

SQL Server upgrades don’t have to be stressful! In 2019, Straight Path’s founder, Mike Walsh, walked folks through the things to consider before, during, and after a SQL Server Upgrade process. You can watch the videos of this series over at our blog at the SQL Server Upgrades Made Easy page. (JON!! LINK)

This document can help you get on the way to having a SQL Server Upgrade checklist. You’ll need to cut and paste from this, add in the details about your own environment, test all the steps to be sure they work right for you and use this as a guide for your own checklist creation process. But please, take all you can from this and use this as the baseline for your own checklist! If you have a good rollback plan, you may physically have a second chance to do your upgrade right! Don’t burn the goodwill of your users and management by doing a SQL Server upgrade without planning!

At Straight Path, we help a lot of folks with SQL Server upgrades. We still double and triple check the basics and rely on checklists for success!
This document won’t answer your questions and do the steps for you! It would never end! It’s to get you thinking of the questions to ask and steps to do! Look to our blog and videos for some thoughts on possible answers to the steps!

For The Lawyers…

NOTE – AS ALWAYS – THIS IS A DOCUMENT FROM THE INTERNET WITH IDEAS FROM OUR TEAM AND BECAUSE YOU WILL BE ATTEMPTING THESE STEPS ON YOUR OWN AND INTERPRETATIONS MAY VARY, THIS IS OFFERED WITHOUT WARRANTY! YOUR MILEAGE MAY VARY AND YOU SHOULD TEST THESE PROCEDURES AND STEPS AND MAKE SURE THEY WORK FOR YOU AND YOUR ENVIRONMENT AND MAKE SURE YOU ARE ADDRESSING ALL RISKS UNIQUE TO YOUR ENVIRONMENT. THIS DOCUMENT IS PROVIDED AS A GUIDE FOR YOU TO CREATE YOUR OWN CHECKLIST! IT COMES FROM OUR KNOWLEDGE AND WISDOM AND THE MANY SUCCESSFUL UPGRADES WE’VE DONE! (Mike’s note – these really are the high level steps and some low level thoughts we use when we help someone do an upgrade)

Search for:
Archives Archives Select Month March 2019  (2) February 2019  (8) January 2019  (2) November 2018  (1) September 2018  (1) June 2018  (2) March 2018  (1) February 2018  (1) January 2018  (1) December 2017  (1) October 2017  (7) September 2017  (2) August 2017  (2) June 2017  (2) May 2017  (2) February 2017  (4) January 2017  (6) December 2016  (1) November 2016  (3) October 2016  (2) August 2016  (2) July 2016  (7) November 2015  (1) October 2015  (1) September 2015  (1) February 2015  (1) December 2014  (1) November 2014  (2) October 2014  (4) September 2014  (5) August 2014  (4) July 2014  (1) December 2013  (2) November 2013  (4) September 2013  (1) March 2013  (3) February 2013  (1) November 2012  (4) September 2012  (1) August 2012  (2) May 2012  (1) April 2012  (1) February 2012  (4) January 2012  (3) December 2011  (2) November 2011  (5) October 2011  (3) September 2011  (1) August 2011  (1) July 2011  (1) June 2011  (4) May 2011  (8) April 2011  (4) March 2011  (3) February 2011  (8) December 2010  (9) November 2010  (8) October 2010  (5) September 2010  (2) July 2010  (10) June 2010  (1) May 2010  (7) April 2010  (7) March 2010  (5) February 2010  (2) December 2009  (12) November 2009  (8) October 2009  (9) September 2009  (2) August 2009  (6) July 2009  (3) June 2009  (5) May 2009  (1) April 2009  (6) March 2009  (5) February 2009  (5) January 2009  (7) December 2008  (5)
Free Consultation

Get help now from experts you can trust.

For I am not ashamed of the gospel, for it is the power of God for salvation to everyone who believes, to the Jew first and also to the Greek.

- Romans 1:16

The post SQL Server Upgrade Checklist appeared first on SQL Server Consulting - Straight Path Solutions.

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

In February, I started a webinar series helping folks planning out their SQL Server Upgrades. The official “Webinar Portion” of this webinar series ended last Thursday when I talked about “Part 4” – thinks to consider AFTER you upgrade. Honestly, these are things you should do whether you upgrade SQL server or not! 

I also blogged last month about a new thing I’ll be doing here. I enjoy teaching and sharing. I’ve been blogging here since 2009, and I’m looking to share more video content. It will always be free, it will be geared towards topics at a 101 or 201 level and I want to keep doing webinar series. Basically, pick a topic, and do a series about it. We’ll eventually share the videos and transcripts here, and perhaps myself and others at Straight Path will write some blog posts that help you out with the topic as well. (Note – We still have some upgrade posts in mind and we’ll get them going. We wrap up this Webinar with a live google hangout to chat about the series, upgrades, SQL Server and whatever else we fancy – or, hopefully, whatever YOU fancy. Check that Google Hangout out tomorrow, Thursday 3/7/2019 at Noon!

While on the youtube page reminding yourself about that webinar, consider subscribing to our YouTube channel – all of our free webinars about all topics will go there for you to watch.

Upgrading SQL Server – Week 4 “After You Upgrade SQL Server”

You can watch the previous 3 weeks on youtube

This is the last video from the webinar portion of the SQL Server upgrade webinar series from the past four weeks. Here we chat about the things to consider AFTER your SQL Server upgrade.  Health checks. Monitoring. Keeping the lights on. Frankly, this webinar is a must whether you are doing a SQL Server upgrade or not! In a week – after our live hangout video goes live, we’ll make a single page that has all of the videos, download links, references to blog posts and other content you might like and if we come up with more free checklists or downloads to help you with your SQL Server upgrade projects, we’ll throw it there and let that page be a living page. Subscribe to the blog or come back in a week to check it out!

After Your SQL Server Upgrade - Upgrading SQL Server Webinar Episode 4 - YouTube

You can also sign up with the webinar ninja series to get the full cuts from Webinar Ninja series and watch replays there and see any Q&A each week including the little crosstalk before and after the “recorded” portion of the webinar starts. Might even see me “dancing” sadly!!

Subscribe to the youtube channel to be alerted of the live chat hangouts and to see the upcoming talks in our continuing webinar series.

Towards the end of March, I’ll start with a multi-part series on High Availability/Disaster Recovery options for SQL Server. If you have questions, you may not need consulting, just some help!  That’s presicely the point of these webinars. Trying to help get some content to you of things to think about when there just isn’t a training budget but you’d like level up a bit. Perhaps in the future we’ll go 200/300/400 level and consider freemium and premium content. For now, though, these webinars will be live, recorded, and free to watch either way.

I hope to see you in the Google Hangout with some of the Straight Path team and I discussing upgrades on the live chat thinga-ma-jiggy at this link.

Downloads from the SQL Server Upgrade Series

(We’ll have a couple more pieces of content still to come on the final recap post that links to all the videos, and all the downloads and plus we’ll work to get links from all the videos when we make that “Central Page” for the webinar series! The upgrade checklist is ready but we’ll put that into it’s own post today, 3/7/2019!)

Transcript

(I’ve used a transcript service for this and I reviewed some of it. It’s a work in progress and there may be some errors!)

Welcome back to the webinar series. Today’s the last actual webinar in the upgrade series, so we’re gonna talk today about things to consider after you upgrade SQL Server. Again, like I’ve said a couple times in the blog post, and like I mentioned ahead of time, this webinar really has little to do with the SQL Server upgrade, but it’s my chance to extol some of my virtues on you when it comes to best practices and configuration and desired state and all the stuff that we should be worrying about but sometimes don’t have time to do. So, like we talked about before, when you’re done with a SQL Server upgrade or when you’re doing a SQL Server upgrade, this may be something that a company might do once every five years, once every ten years as some people here would say, and if we’re doing it so infrequently, what better thing to do than to use this as an opportunity to start fresh.

Pretend it’s like New Year’s Day for your SQL Server. You’re gonna finally eat better and workout or whatever your resolution is, except for this is gonna be much more serious than New Year’s resolution because you’re actually gonna do it, so when you’re done, how do we know we have a best practice to set up on our SQL Server? How do we know our SQL Servers are set up for success? And, that’s what we’re going to talk about today.

If you’re new to the series, this is where we have been. We’ve done a lot of things. We had a big picture at the very beginning. That was more just talking about why, why, why, so why are we going to upgrade? What’s the point?

Then, we talked about what do you do before you upgrade? How do you plan it? How do you ask the right questions and get the right people involved? We talked about checklists, and that’s one download I’m still working on finalizing for you, so that’ll be in the recap. We do have one PDF we’ll talk about in a second, to download today. From last week, the SQL Server health and best practice checklist, we finally released that, so you could download that on a blog post that went up today.

Then today, we talked last week about the upgrade, so how do you actually upgrade? How do you actually go in there and do your upgrade, we talked about … Again, in 30 minutes, I’m not gonna tell you how to upgrade your SQL Server from start to finish, but we talked about the important concepts.

I think the HA/DR web series might be a little more involved. We can stretch it out a bit and maybe do a few more weeks in there and actually do some demos, and AGs and FCIs and even log shipping. But this, I don’t wanna get into demos because there are so many. I’ll post the link. Actually, I’ll post it right now. If you just go to my blog. Just go to my blog, Jose, and you’ll see the latest post is the recap from week three.

In there, there’s two links. There’s only two downloads right now. There’s the 20-ish page SQL Server Configuration Guide, and then there’s the much smaller PDF of the Do Your Own Health Check guide. That’s cool, but the actual SQL Server Configuration checklist is a recap of all of our best practices that we look through when we do a SQL Server upgrade. Checklist is probably a strong word for it. It’s really a document with bullet points, but I’m sure we can probably turn it into a checklist if people would like it that way.

Then, the last thing I’m working on is an actual upgrade checklist. Again, I proposed … I might just make an Excel spreadsheet and post that instead of trying to make it PDF and make it marketing friendly. That’s the hang up here, is trying to get into PDF, but I’ll figure out the right format, and I’ll put that, when I share I this video from this series next week, I’ll share that.

Next week, I’m not traveling as much, so I should actually have time to get the blog post out by probably Monday or Tuesday, like I hope, and get the transcript servers to run sooner. Anyway, that’ll be the last downloadable thing we have there, is the upgrade checklist, or the before and after and steps there.

Anyway, we’re here today, we’re going to talk about after. We’re going to talk about what do we consider when we’re all done upgrading, you’re not just done when your upgrade is done. And if you think you’re done when your upgrade is done, you’re lying to yourself. You should never be done with SQL Server. We’ll talk about that in a second.

Then next week, we’re going to do a live hangout. So, this will not be on WebinarNinja. You have to go to that link right there, that YouTube.com watch and that will take you to, I think, to my channel, the Straight Path, our channel I should say, the Straight Path channel. Once you’re there, there’s a little remind me or whatever icon, you click that icon and Google will somehow let you know, maybe give you an iCal to remind you for the live hangout. Next week, same time, we’ll have the whole Straight Path technical team, or a good portion of us anyway. I know Tara will be there. I know I’ll be there. I think Mike Clark will be there, I’m not sure. And maybe Jack and Joe, and maybe Bruce. So, we’ll have the whole technical team there to answer your questions.

If you have questions, we’ve gotten a couple, if you have more questions, whatever they are, if it’s about a Sequel upgrade or even about the health checks and the best practice stuff, bring the question with you live. Or if you’re too shy or you can’t be live, we’ll record it, just send me an email to mike@straightpathsequel.com and we will answer that question on that hangout. We’ll record it, and when this whole series is all done next week, when I do that hangout, I’ll take the video from that and all these other videos and blog posts, and we’ll make one big, sort of like a spoke or hub blog post that lists all the videos in order with more download links. And as I re-watch the videos and listen to my annoying voice, I say, “Oh, I should probably add that,” so I’ll add a lot more extra links that nobody asked about but that come to mind while I watch it. So, that last post will hopefully be a guide and maybe even have some checklist steps in there.

The goal again, is not to upgrade your SQL Servers in five 30-minute webinars. It’s really to get you thinking about the right things, and to help get you your questions answered and to put you on the right track.

So, thank you for coming back, by the way. A lot of you came back and I appreciate that. You stuck with us, so either you have nothing else to do or you appreciate the content.

Who am I? You all know by now. Most of you know who I am. I’m Mike Walsh. I run Straight Path Solutions. We’re a SQL Server consultancy. I’m not doing this as a SQL Server consultancy thing, I’m doing this really as an MVP thing. I’m just doing this as a person in the community trying to give back. Whether I keep my MVP or don’t, I don’t care, I just like sharing. Our company does a whole bunch of things. But one of the things we do a lot of, being a consultant company, is come alongside people and do upgrades.

As a result, I’ve been burned by a lot of upgrades, and I’ve taken over upgrades that they didn’t think about these things ahead of time. They didn’t go through a checklist. They didn’t do best practices. We’ve been there to catch people when they fall from upgrades not well thought out. I want to avoid that for you. I want you to avoid that.

Where are we? Again, there’s some recaps if you go to straightpathsequel.com/blog you’ll see pretty much all the latest posts there. There’s one availability group post in there. But, the other latest posts are all about upgrades. I have a couple more posts in the queue, in the draft mode, that will be about upgrades as well, and again, we’ll have the big recap post next week with a bunch more links. Really, the content for the rest of the next week or two on my blog will be about upgrades. I’m going to switch over and make it HA/DR stuff, because that will be the next webinar in March.

Again, the same idea. We’re not going to go in depth. We’re not going to be 500 level material. I want to start at the beginning and what should you think about before. Which approach, how do you do RPO and RTOs, and how do you balance budget, and can AGs do this? Can FCI do this? How do we mix and match, and how can we do it on a shoestring and what shouldn’t we do on a shoestring, and all these things. I want to talk about before, during and after. In fact, every single webinar we do, I really want to hammer this whole idea of things to think about and do before, things to think about and do during, things to think about and do after. I think if we simplify it to that, we can make simple. Right? That’s my hope.

The next series will be probably two weeks after the live hangout, so towards the end of March. That will be on, again on High Ability DR, so the content will start switching over the HA/DR at that point, I think.

By now, we should know all the things we knew last week, the week before. We should know why we want to upgrade. We should know what we need to think about. What questions we should be asking. You know, who, what, when, where, and why, from two weeks ago video. We should really be knowing where we want to go, how we want to get there, and really why we want to go.

Now, we should also know how because last week we talked about how. If you all remember one thing from last week, I don’t like in place upgrades, and really frankly, you shouldn’t either. It doesn’t save you time and the rollback can be bad. If you have no choice, you have to be careful and plan it out, and spend extra time making that rollback checklist in the go, no go criteria. But you should have a better idea of how you want to upgrade. How you’ll test it. Who’s in charge. That sounds so simple but so many times you can have conversations and somebody thinks that they just got assigned a task … or they think they assigned you a task but nobody knows because nobody’s just done that little … here’s the plan, here’s your role, here’s your role, he’s your role, and people who asked before, I forget who it was now, maybe Dale, you know, what if four weeks ago, you have all the hats? You’re the DBA, the SysAdmin, or again, well have the conversation with yourself again to understand what the right roles are and what the right pieces are.

Then, we’ll talk about the rollback. Talk about the steps, and that’s what we should know from the before, and during upgrades. That’s it. We’re upgraded. Congratulations. The webinar series is over, right? No. No. Remember we talked last week about this. Y’all remember what this is, right? You don’t have to all type in chat but it takes hard work to type a few keys, but we know what this is. This is the Show Time rotisserie grill. Actually, I think my in-laws have one and it actually works okay. You can cook Italian sausages in it, by the way. But I’m not [Roan Po-peel 00:10:07]. It’s really good if you don’t like greasy Italian sausages in spaghetti. But anyway, I really digress now.

This is a chicken making machine. It’s set it and forget it, is what he had all the people in the infomercial audience saying over and over. “Set it and forget it.” This is not the SQL Server rotisserie chicken machine. This slide right here, while I’ve the changed the SQL Server logo, the very first talk I ever gave on SQL Server, one of my slides was the rotisserie chicken machine with a SQL Server 2008 logo on it. Probably 11 years ago. I had this presentation I gave called, “So You Want to be a DBA? Where Do You Start?”

So many people, not you because you’re here at the webinar and you care about the Sequel community and you care about learning, but so many people think that SQL Server can be set and forgotten. Right? You install it really easy. There’s a few more clicks to have to do, but for the most part, it’s next, next, next, next, finish. Then you can go away, have some of your, no brand name, have some of your coffee, a little sugar. It’s Americano. But you can just walk away and come back and poof, you have a SQL Server. Then, you can take your installer from your client, for your ERP system, or whatever application you’re using to do critical business, and you can install that application and it will create it’s own database, for the most part, or maybe you have to create it. Then you can walk away and you have a SQL Server.

I can’t tell you how many health checks I do for clients, big and small clients, where it’s like, “Oh, you don’t take backups? Did you know you don’t take backups?” “You don’t do check DBs. You don’t do this.” I see it all the time. Not you. None of you do that. You’re watching this video, your here at this webinar. You’re not doing that because you care about best practices, but other people do.

SQL Server, and I’m not a slide guy so you can make fun of my animations all you want, but there you go. That’s the extent of my animations. Rotisserie chicken, SQL Server, no. A bouncing no, too. And a misshape, but SQL Server’s not set it and forget it. That’ the main point of this one webinar. We’re going to talk about just again, things to keep in mind.

This almost goes back to one of the before webinars where I told you that you have to do a health check first. If you were obedient and you did your health check before, you’re going to be okay, because you saw what was bad and you care, and you made a list, a checklist even, of what you want to do new and better on the new server and you fix it. Right? Didn’t you? I hope you did. I hope you’re going to.

If you didn’t do the health check and you’ve upgraded and you did a set it and forget it upgrade where you just do your upgrade and your done, well, it’s not too late. You can do a health check. In fact, I recommend you taking one of the free health tools out there from all the people in the Sequel community who worked hard to build them, and do a SQL Server health check once a year. You know, does DBA checks, [inaudible 00:13:21] I don’t mention here and in fact I actually, I’m going to make a note here.

Normally, our office manager comes to the meetings to take notes but she’s actually on a client call. Some of us have to do work. So I’m just going to make a note here in my notebook. Give me a second.

I’ll make sure I include some links to other tools. I mentioned Glen’s script. I mentioned Brent’s blitz scripts, but there’s so many other ones. And I’ll mention a few others. Try a different flavor. But, once a quarter, once a half year, once a year, do yourself a favor and give yourself a SQL Server health check.

Why, Mike? The configs don’t change. I know they don’t change but new databases got created. New users got created. New logins got created. All things have changed. Maybe not your max [dop 00:14:01] and your max memory. Hopefully that’s not changing as you go, but it could. You could have other people who are helpful DBAs in there, or vendors with the wrong level of access because you didn’t fix security, undoing your best practices.

So, after you upgrade, and if you never upgrade, and if you go to the cloud you never have to worry about upgrades, do health checks. Make sure your server’s good. Seriously, it saves a lot.

Dale is saying, great point Dale, he says, “Applications change and new store procedures get installed. They may not perform as expected,” and there are some client’s tools out there that will actually change your system configuration because the vendor thinks they know best and they need SA for the installer, and the installer will actually change some settings. It’ll turn XP command shell on, and then when you get audited in nine months, the auditor says, “Why is XP command shell on?” And you say, “It’s not. I turned it off.” “No, it’s on.” And you say, “I don’t know who did it,” and you turn it off and then you get a phone call from accounting saying that the system’s not working.

You can not just install and walk away. The DBA, whether that’s your official title or not, is the person at the company, again I’m being, this is hyperbole, sort of. But you as the DBA, I think it stands for Database Advocate, you are the one who cares about the data. In fact, I’ll say … and you can write me a letter if you like, you’re the only one who cares about the data in the company. That may not be true, but you should act like it’s true. Act like your the only one who cares about the data, and treat it well. That means do health check, that means when that bad application vendor tries to do something naughty, you get mad at them and you say something about it.

Some best practices evolve, right? If you were to get a SQL Server health check from me, 10 years ago, before … I think it was before Glen started writing his diagnostic scripts, maybe about the same time, but anyway, 10 years ago before I had a blitz or Glen’s scripts and I had my own …

I had a blitz subscription. I kinda had my own. I bet you there are things that I would tell you about today, that I didn’t check 10 years ago. There are things that I forget about today, or that I find today, that I didn’t even think of three years ago. So some best practices evolve and it’s okay to learn but the trick is you have to keep learning. Right? So towards the bottom in italics there, I say, “Always learn.”

Part of what you do after your upgrade is don’t just move on to other things and forget about your basic PBA skills. Go to classes, take webinars, come to the free HA/DR webinar. Maybe we’ll do one on SQL best practices and performance tuning and configuration. Learn about SARGable queries.

I know we should all know this but I keep finding the same problems over and over so somewhere, some place, people aren’t learning. And, so, if we learn and we do more, we’re good.

So I’m reading from the bottom up. But do the health check. Again, our PDF is a real basic step. So over the two downloads we have, the PDF. One of them is..

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

A few weeks ago, I started a webinar series helping folks planning out their SQL Server Upgrades. (It’s not too late to register to join live and participate in the Q&A and get access to the downloads and any Q&As not recorded)

I also blogged last week about a new thing I’ll be doing here. I enjoy teaching and sharing. I’ve been blogging here since 2009, and I’m looking to share more video content. It will always be free, it will be geared towards topics at a 101 or 201 level and I want to keep doing webinar series. Basically, pick a topic, and do a series about it. We’ll eventually share the videos and transcripts here, and perhaps myself and others at Straight Path will write some blog posts that help you out with the topic as well. Check out the UpgradeSeries category for more of those posts as we continue the series over the next few weeks. Each series will also conclude with a live Youtube hangout with anyone on the team or colleagues in the industry and I chatting about that topic and answering your questions. Feel free to hit that link and let us know what topics you’d love to see a free, interactive, multi-part 101/201 series about.

Upgrading SQL Server – Week 3 “During Your Upgrade SQL Server”

You can watch week one here on the blog or directly on youtube. You can also check out week 2 on youtube or blog. Please consider subscribing to the channel as we’ll make it a rule that all webinar content from the free webinar series will be shared on youtube and sometimes here on the blog also.

This is the third video from the series from the webinar last week. Here we chat about the things to consider DURING your SQL Server upgrade.  We discuss questions you should be asking yourself before any SQL upgrade project starts. We talk about rollback plans, CNAMEs, checking health before, etc.

Upgrading SQL Server Webinar Week 3 - YouTube

Sign up for the to get the full cuts from Webinar Ninja with any Q&A each week including the replays of the past webinars. Subscribe to the youtube channel to be alerted of the live chat hangouts and to see the upcoming talks in our continuing webinar series. Towards the end of March, I’ll start with a multi-part series on High Availability/Disaster Recovery options for SQL Server. If you have questions, you may not need consulting, just some help! Check out the series, attend live and ask your questions or come to the Q&As. I’ll try and answer most questions left in the posts also. Yes, I own a consultancy, and if you need in-depth help, I think we can help – and I trust you can figure out how to reach us. But the point of this content and the webinar series is to help encourage you and show you how can you do more than you realize! See you on Thursday 2/28 for Week 4 – “After you upgrade”. Then on 3/7 we’ll do a live hangout on Youtube with the Straight Path technical team answering your questions.

Downloads from the Series

(The upgrade checklist download will come out in a separate blog post by itself next week. And after the Live Hangout on Youtube, we’ll take all the videos, all the content and make a single page with all the content)

Transcript

(I’ve used a transcript service for this and I reviewed most of it. It’s a work in progress and there may still be some errors!)

So today, we’re going to talk about doing the upgrade. Today we’re going to talk about … we know what to do before, we know why. What are the steps we should do to actually have a successful upgrade? And what are some pitfalls we can avoid? I need to get over here because my slides are being controlled differently, apologize. Let’s see if I can do this right. Perfect, I can. So, again, this series, we’re really … Oops, yeah, I did it too fast.

We’re in week three, and here in week three … let me see if I can move my mouse right on top. There we go. Here in week three, we’re talking about doing the upgrade. So, some best practices for the new SQL install. How can you minimize downtime, right? How can we get the environment upgraded with having our users waiting nervously for less time? How can we do it with success? What are some ways to test to know we got all the data?

Next week we’ll talk about after the upgrade. Really, next week has nothing to do with upgrades. It does because we should be looking at these things after we upgrade, but if you don’t upgrade your SQL server and if you’re not planning on an upgrade, next week’s going to be packed with good content for you to just be a better DBA and to watch your environment better. So who am I? I’ve said this each week. My name is Mike. You can find out more about me on my blog, StraightPathSQL.com. I run a company called Straight Path. We do SQL server consulting, manage services, we help people upgrade projects. We do all sorts of SQL work.

I have a lot of upgrade scars. Being a consultant, we get to do a lot of upgrades, right? So when I was a full-time DBA, I could probably count on one or two hands the number of times I had to upgrade a SQL server. The only reason I have two hands is because I was a job hopper early on. I was always a consultant, I just didn’t realize it. Most people just don’t do a lot of upgrades, but we do a lot of them as consultants because we have a lot of different clients at different phases of the life cycle. So we’ve seen the problems, and we’ve gotten called into the upgrades that went, oops, or we’ve had the people who didn’t heed advice and tried to upgrade a path that maybe we don’t like, and got burned by it.

So where are we? Again, I already kind of gave the recap. These are the things we should know though. After the earlier two webinars, we should know why we want to upgrade. Obviously, that’s a pretty easy question, but we should know some of the reasons why we want to and some of the reasons why we can … the reasons that maybe our management wants to hear. So, hey dear leader, here’s why you should want me to upgrade. So, that little reverse psychology. Then we’re going to … We already talked about where we want to go, so we should have an idea of that. Again, you need to do the upgrade analysis with the upgrade advisor. We talked last week about the DEA tool, the database experiment analysis tool. We can talk about doing a SQL server database tools last week and building towards your target. The point is, we should know roughly where we want to be.

We should have an idea of what our environment looks like. We talked last week about running a health check and I have you guys … talked about link, and that link for the download is actually on the blog post from week two’s video already. We can run our health check PDF, or you can go and download the blit scripts, the Glenn Berrys scripts, or any scripts. I don’t care. That’s well trusted by the community, and see where you are. We should have a performance baseline, so we make sure that the new environment we build will have the right performance.

We should have the rough idea of which approaches we want to use to test. How do we know that we can support the new capability mode, because going to the new version is really cool, but it’s really important to also be able to work in the latest compatibility mode if we can so we can take advantage of the new features. We can use [inaudible 00:03:50] functions for example. Then we should have an idea … We talked last week about role back and sort of go/no go criteria. I said, it’s not a go go. It’s just go/no go. Most companies don’t have a no go. They just have a go and we’re going to go whatever you say. But we need to be the pesky annoying DBAs who say, yeah but, and explain the problems. So, that’s kind of there we should.

So where are we going to go? Well, today is going to be about how. So we’re going to zoom in on the how. How should we install SQL? We’re not going to go into in depth on that because that’s a whole webinar by itself, but we have a download we’ll have in the blog post about this for that. Like I said, I can’t give you the links to webinar ninja. The people watching the videos won’t be able to get them anyway, so we’re going to put them on the blog post. What’s going to happen is we’ll do a webinar on Thursday, and then by Tuesday of the next week … so three business days later, there’ll be a blog post with a video and a transcripts and links we talked about, and any downloads that we say we’ll make for you.

Normally this downloadable content, we’ll probably use a lead magnet on our website and people can come and give us their email address and all that. But for people coming to the webinar, I just want it to be a free thing for you, so you’ll just have a direct download link. Get it. I don’t want your email address. Just take it and use it. So we’ll have a checklist that we use to help guide us on SQL sort of best practices. That checklist alone could probably be a three part webinar series, but we’ll give you that.

How should you organize your own checklist for the upgrade? We’re going to talk about that today quite a lot. Then again, next week on the Tuesday, we’ll have a sample checklist that you can use and fill in. We’re going to talk about how to best execute an in place upgrade versus a migration based upgrade. How do we minimize downtime? How do we do a test ahead of time? How do we do a test after? These are the things we’re going to talk about this week.

So let’s first … I just want to talk about in place upgrades first, get this out of the way. I have a pretty simple in place upgrade process. My first step for in place upgrade is you have to attain the SQL server installation media. So go to MSDN, go to your volume licensing portal, wherever you get it. You can even get the old DVD. So this is SQL server 2012. At the [inaudible 00:06:03] they gave us a pretty cool sign book. The DVD’s not in there. It’s just an empty package with all the signatures of the program team.

It’s kind of neat. I keep it up here. It’s an old version of SQL and I don’t encourage you to use old versions of SQL, but I just like this little … I don’t know, it’s nice. It’s nice. I like nice stuff. It’s just touching. Mass produced, but still touching. So then you run set up. So you get your install media, you get it up there, you run set up executable, and then you stop. Seriously, you stop, throw it away. Don’t do an in place upgrade. If you were here hanging on the edge, just like oh this is great … I understand there are situations in life where you have no choice. You’re on a physical server. The management says, this is how we’re going to do our upgrade and we’re not buying a new server and we have no swing server, deal with it.

Yes, there are situations where an in place upgrade is your only choice. I’ve done them. I’ve helped people do them. They are so risky. I don’t like in place upgrades. As far as it’s up to you, put your foot down and say, we will never do an in place upgrade. Here’s the deal … if you’re on a SQL server 2008, when was it last built? If you happen to buy brand new SQL server hardware last year and you just moved your 10 year old SQL server 2008 to it, well yeah it’s new hardware. Maybe they don’t want to spend the money, and I understand that. But most of us are in virtual. Many of us are in the cloud. There’s other ways to do it.

We had one person ask about in place upgrade and we said, we don’t recommend it. Here’s all the reasons why, even though it’s simple. They went ahead and tried it and they said, nope don’t worry, we’re going to do it. We said okay, we’re support, you’re going to need it. 99%, maybe 95%, maybe 90, whatever, some percentage of the time. Not as good as 99.9, but some percentage of the time, they work flawlessly. When they go bad though, they go really bad. When they go bad in the middle of your upgrade, you have no rollback. Your rollback is to start mashing buttons to hope the old version works to try the uninstaller, and to do a bunch of stuff to your production SQL server that’s been your bread and butter database server for how many years. So I’m not of it. It’s pretty clear.

Again, they’re not that bad. They have to happen sometimes. If you want to see more, I’ll have a link for this. But if you just do a google search with DBA.stackexchange.com and you search for in place upgrade, you’ll find this answer I gave. It’s well uploaded. I have 91 votes on it. This is just one page. You’d have to press the page down button like four times. Anybody who’s in this webinar whose worked with me as an employer, as an employee, as a customer, you’ve sadly probably seen my long emails. I’m working on it, but some of my stack overflow answers can be long too.

My short answer is no, don’t do in place. If you have a question about in place upgrade, throw it in the comments and we can talk about it. We can talk about the live webinar, but we [inaudible 00:09:15] the Q and A in a couple weeks.

So the first thing we do is we’re going to install SQL server on the new. Hopefully you’ve actually done this in test already a couple times. We talked about this last week. You want to test, test your migration, test your compatibility mode, have your applications point to it, get a test version of your … Ideally, the ideal thing to do is for you, before you do your upgrade, to do an upgrade advisor, maybe to do the DEA tool, then maybe take SQL server database tools and run a build against it. Now that’s for your database. If it’s your vendor database, it might be as easy as talking to your vendor and ask them what do they support, and just cross your fingers, pray, meditate. Whatever you need to do, hope that your vendor is going to be compliant. Oh yeah, of course, we’ve already tested it and we certified it. You’re good it’s easy.

Sadly, a lot of our vendors just don’t play that game for us. Actually, I shouldn’t say a lot. Half do, half don’t. So it’s 50/50. But if it’s your own application, you want to actually run a build against it, against the target for the version you want to go and just see what breaks. But all of those tests alone are not as good as deploying SQL server X. So let’s say you’re on 2008 today and you want to go to 2017, upgrade to 2017 and test, deploy your database, get it all good there, and then point your test suite to it. Point your applications to it and do a full regression test and see what breaks. There is no better testing than a full regression test to see what breaks.

Once you’re done, build a new server. Build it right. So remember, you have an awesome opportunity here. You have a chance to take all the knowledge you’ve gained over the past 10 years and apply it to this new server. You can [inaudible 00:11:00], you can make sure power saving is enabled, you can finally have 64K drive allocation size when you format your drives. You can, if you’re on a VM, you can do the right data stores and para virtual scuzzy card connectors, and you can do all the best practices that you keep reading about in the SQL server community, but you say well I don’t want to touch it because it’s already here and we don’t want to break it. We want to wait for the next version. The next version is here.

You have no choice, by July, if you’re on SLQ server 2008 and you don’t want to spend a bunch of money on Microsoft and move to Azure and get their secret slug worth deal, or whatever his name is from Willy Wonka. But you have a chance now to build your SQL server to best practices. Run a health check on it, even though there’s nothing on it yet. Run a health check and see what settings you got wrong. Add maintenance. If you don’t have all the [inaudible 00:11:53] maintenance scripts or some trusted maintenance solution, install that in the new server and experiment with it. So when you migrate over there, you don’t have to say oh yeah … because here’s the deal, tomorrow is a big enemy of us. It’s an enemy of me. Tomorrow is the worst word in our career.

How many times have migration happened, and it finally gets done, and something goes wrong, a little bit minor but it’s not big. You get it solved at 1:00 in the morning and you say, alright, in the morning I’ll fix the backups. In the morning I’ll fix the index maintenance. In the morning I’ll fix this. Why do that? Set up your maintenance the right way today and have it scheduled automatically. It’ll be really quick right now when no user database is on it for the two weeks you’re waiting to move it migrate, but set it up head of time and then it’s just good to go. Set it up automatic is good. Setting up ahead of time is good.

Don’t recreate security woes. Does the whole entire world have SA access to your SQL server? If they do, fix it. Now, there’s two schools of thought there, I get that. Some people will say, well yeah, but we’re doing a migration. We want to break as few things as possible. I get that, sure. But if you can start over and have the best practices in mind, do that. Again, if we have a test first mentality, we’ll test this and we’ll see what breaks. Does your vendor really need SA access to your consolidated SQL server? Probably not. Will your compliance people like to know they do? Probably … Not even probably. They don’t, period, end of story. They don’t even like that you have SA access.

Again, you want to pick the right resource. So we talked last week about grabbing some perf data and doing the baseline analysis. If we do the baseline analysis and we know what we need, and we know what we’re using, we can actually build the new server to spec, and build it with the right amount of resources. Yeah, so Mark just mentioned a great comment. Mark says, “I have nothing but praise for Ola Hellengren, but make sure that’s his default are appropriate in your environment.” Yeah, absolutely.

We just had a conversation the other day with somebody on my team about this. Not somebody on my team, but somebody off my team. The defaults are bad. If you just take the default in the next maintenance job and run it, you’re doing the next maintenance. If you schedule it every night, like you just want to update your status, every night you’re doing an index rebuild and you’re not even updating your stats. So don’t just … True of any script and run it. You want to actually go to Ola.Hellengren.com, or whatever his website is. We just do a google search for Ola SQL and you’ll find him, but you can say Ola Hallengren, H-A double E-E-N-G-R-E-N and SQL, and you’ll find his blog, or his maintenance [inaudible 00:14:39] website really.

If you just click in any of the links for index optimize or backup, you’ll see pages of options. He even gives examples, so there’s really no excuse to run the defaults. Do not do it. Thanks Mark, that was a great point. Don’t forget to enable the maintenance clean up. So you have a new server. Keep your MSDB happy from the beginning. Do your MSDB cleanup history right away. Do your job agent history cleanup right away. Keep these things pruned and well oiled from the beginning. Recycle your air log on a regular basis. We’ll talk next week about some best practices in the after and kind of talk about some of the things that we find that health checks commonly, and getting your monitoring set up again. We’ll talk more about that next week.

Here’s a big thing. I’m a big fan of CNAMEs and aliases and not connecting to server names. If I have 10 applications on a consolidated SQL server, I might be as crazy as asking IT for 10 CNAMEs. Now you don’t have to go that far if you don’t want to, but I don’t like people connecting to the SQL by its actual host name. I want them to use a CNAME always. There are schools of thought. Is it silly to have CNAME if it’s going to an AG listener, or if you have a cluster incidence, because that’s kind of a C name going to a CNAME. Yeah, maybe, but I like them still. This is because I’m thinking my future upgrade, right?

SQL server is now coming out every couple years. Their servicing model is different. Staring 2019, there’ll be no more service packs. We’ll just have CUs and slip stream CU updates. So SQL updates come fast and furious now. You don’t have to take the train every time it comes by your stop, but you shouldn’t be four trains behind, like many people are today, or five. So, if we’re going to start upgrading more frequently, we want to make our upgrades easier. If we take the pain today and make a CNAME and migrate to it, that does two things for us. We’ll talk in a second when I talk about operation honey pot and how that can help us, but it also makes future migrations easier.

So you’re making today’s self mad. This is stupid. But in two years, your future self will say thank you. I could sit here for hours and talk about all the ways that old Mike screwed over today’s Mike and I hate it. Right here at the farm, we’re running out of our fire wood. We probably burned five cords this year. It’s been so cold a couple times and we’re not using oil. I cut my wood way too late, so it’s way too green and I’m probably going to buy a cord of wood this weekend and it’s going to be way too expensive, but I’m stubborn and I don’t want to burn oil.

I’m mad at the Mike who was sitting on the couch last spring or two springs ago instead of cutting down woods here at the farm. Take the time today to make your future self like you. It will be amazing. So make the CNAME. When we talk about operation honey pot in just a few slides, it will give you another tool to find those non-compliant, frustrating developers and vendors. So we’re not going to read all the slides here. You can get the slides on the blog post when I share it next Tuesday, but it starts with a plan, right?

So you’ve got to have an upgrade checklist. I love checklists. I don’t use checklists nearly as much as I should. I’m on the fire department. We have some checklists there for truck checks, and we have checklists in the ambulance for certain cardiac emergencies...

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

A few weeks ago, I started a webinar series helping folks planning on a SQL Server Upgrade. (It’s not too late to register to join live and participate in the Q&A and get access to the downloads and any Q&As not recorded)

I also blogged last week about a new thing I’ll be doing here. I enjoy teaching and sharing. I’ve been blogging here since 2009, and I’m looking to share more video content. It will always be free, it will be geared towards topics at a 101 or 201 level and I want to keep doing webinar series. Basically, pick a topic, and do a series about it. We’ll eventually share the videos and transcripts here, and perhaps myself and others at Straight Path will write some blog posts that help you out with the topic as well. Check out the UpgradeSeries category for more of those posts as we continue the series over the next few weeks. Each series will also conclude with a live Youtube hangout with anyone on the team or colleagues in the industry and I chatting about that topic and answering your questions. Feel free to hit that link and let us know what topics you’d love to see a free, interactive, multi-part 101/201 series about.

Upgrading SQL Server – Week 2 “Before you Upgrade SQL Server”

You can watch week one here on the blog or directly on youtube. Please consider subscribing to the channel as we’ll make it a rule that all webinar content from the free webinar series will be shared on youtube and sometimes here on the blog also.

This is the second video from the series from the webinar last week. Here we chat about the things to consider BEFORE you start your SQL Server upgrades.  We discuss questions you should be asking yourself before any SQL upgrade project starts. We talk about rollback plans, CNAMEs, checking health before, etc.

SQL Server Upgrade Webinar - Week 2 (Before you Upgrade Your SQL Server) - YouTube

Sign up for the rest of the series and to get the full cuts from Webinar Ninja with any Q&A each week. Subscribe to the youtube channel to be alerted of the live chat hangouts and to see the upcoming talks in our continuing webinar series. In March, I’ll start with a multi-part series on High Availability/Disaster Recovery options for SQL Server. If you have questions, you may not need consulting. Just check out the series, attend live and ask your questions or come to the Q&As. I’ll try and answer most questions left in the posts also. Yes, I own a consultancy, and if you need in-depth help, I think we can help – and I trust you can figure out how to reach us. But the point of this content and the webinar series is to help encourage you and show you how can you do more than you realize! See you on Thursday 2/21 for part 3 on “Doing your SQL Server Upgrade”!

Links Discussed Transcript

(I’ve used a transcript service for this and I reviewed most of it. It’s a work in progress and there may still be some errors!)

Welcome back. Last week we sort of talked about why should you upgrade and the real answer is “why not upgrade?” I mean SQL Server has changed so much since the SQL Server 2008 that you’re running or the SQL Server 2012 or 2014 even and hopefully not the SQL 2005 and 2000, hopefully you’re off that. So I think we all understand the why, it’s really why not. But the world has definitely changed as we talked about last week.

So what I want to talk about this week is a little bit more about “the before.” What are some things to consider? And this is going to be not demos, it’s really going to be chats so if there’s questions ask them live in the chat, if it’s germane to this topic we’ll kind of put me off topic and we’ll talk about it. We’ll keep this 30-35 minutes. I have a little more slides than maybe that time so if we need to we might go longer or I might just cut myself off and extend more next week.

So anyway we’ll go in here. So again this is the series. Last week was big picture, kind of a “why why why” and what’s changed. And I gave you some homework to basically analyze the versions that you think you want to be at, analyze why you wanna upgrade, maybe start conversations with your vendors because the vendors are the pains here and they’re the ones preventing you from upgrading in a lot of cases, talk to the company and really see what are you going to do. I mean especially if you’re on SQL Server 2008, 10, 11, 12 year old database technology that’s gonna be end of life, with all those caveats we talked about last week where Microsoft is making some special deals. If you just give them more money they’ll still give you some patches or if you go to Azure they’ll give you the patches for free. And I talked to you about my opinion of that. So it’s time to upgrade. And the homework was really to start getting yourself ready and getting your organization ready.

Today we’re gonna talk about what to do before, how do we plan our upgrade project right. We’re about to do a massive change to our SQL server. It’s not easy and it’s not small taking a SQL Server that’s running really well here for 10, 11, 12, 14, even 7 or 4 years, and move it over here to a totally different SQL Server version, to a completely different infrastructure potentially, upgrading to a version that you’ve never been on. The world’s changed. You’ve been so busy you’ve not learned new stuff. So what are some things we can do to make this successful and that’s what we’re talking about.

Again I’m giving you something that could probably be a 30 or 30 hour course and a lot of consulting. I’m trying to give it to you in a handful of 30 minute webinars. So we’re not going to cover everything but we’ll get to a lot. I already told you who I am. I’m a guy who’s a SQL MVP, I’ve been doing SQL for 20 years, I’ve been consulting for 10. I run a company called Straight Path Solutions. And more importantly that’s my blog. I’ve been blogging there before I was consulting there and so the company name just formed after my blog name. I’m a user group leader and I have a lot of scars from SQL Server upgrades. In fact I used to have a full head of hair before I started upgrading SQL Servers.

So this is a good quote from a poem. It’s, “The best laid schemes of mice and men” and then it’s said in like Old English. But it’s “The best laid schemes of mice and men often go awry.” So we need to be prepared. Louis Pasteur says “Fortune favors a prepared mind.” When we sit down to begin any upgrade project we want to make sure that we can answer a bunch of questions.

And what are the questions? Well, half of SQL Server or more is just basic human nature stuff. So can you answer the who, what, when, where, why and how questions. And if you can answer those questions you’re halfway to having a good upgrade. A lot of times when we go in as consultants and help somebody do an upgrade it’s not really going in and doing a bunch of really insane and in-depth analysis. Yes, we use our SQL Server knowledge but a lot of times it’s really half sort of what I call “business therapy”. We’re just helping them ask these questions. We’re just coaching them along on saying “What are you trying to do, where are you trying to go, why? Who’s in charge? What team do you need to have in place? What are you gonna go to? So what version of SQL are you going to go to? What’s the problem you’re trying to solve.” That’s also the why question and I have a capital Y in the W slide, I’ll have to fix that later. “When do you want to go there and when are you gonna test it and how much time can you be down for? How can you do it with minimizing that? What are the risks? How are you going to do it?”

Ron says, “Why are you still on SQL Server 2000?” Yes, we’ll talk about that in a couple slides. But absolutely, “How do you roll back.” And next week we’ll talk about it.

I absolutely hate, and I’ll use that word, I hate in place upgrades. In place upgrades when they work are OK and it’s like “Phew, it worked.” But even when they work we’re still on the hardware we were on. We’re on the same OS we were on. We’re still living on a bed of the stuff that we did five years ago or 10 years ago, or in Ron’s case, 19 years ago. We’re still, we’re living on that bed and if we learned anything and hopefully you’re learning something every year … I mean I wish I could go back to a company I used to work at full time and say “Oh I wouldn’t have built like this, I wouldn’t have made the config like this, I wouldn’t have done the operating system like this. I would have done 64K drive allocations instead of the default.” Because 15 years ago I didn’t know about that. I would have built with the CPU configured this way. I would have had less CPUs with better performance and I would have potentially had and standard editions of Enterprise. All these mistakes your current environment is built upon. And they’re not mistakes that are bad per se, it’s not that you’re a bad person because of them, it’s just you didn’t know.

So you know much more now. You’re going to webinars with some guy from New Hampshire. You’re in the SQL Server community. You’re taking classes. You’re reading books. You know more this year than you knew last year. As a consultant I do that all time. I’ll go back to clients, “Hey listen, remember that thing I told you do three years ago? That was good three years ago but we can do it better now and here’s how and here’s why I learned it.” So even if an in place upgrade succeeds I don’t like it for that reason. But when in-place upgrades go bad and “best laid schemes of mice and men often go awry.” Well the best laid plans of in place upgrades can go awry and when they go awry they go awry bad and the rollback, that’s not a fun weekend. It’s like an “I hate my job. I’m updating my resume” kind of weekend. So we’ll talk next week about why I prefer migration style upgrades. But the biggest reason is rollback is just as easy as changing your CNAMES and we’ll talk about CNAMES in a few slides actually.

But it’s as easy as changing your CNAMES, pointing back to the old server and saying ‘I don’t know what happened. We had a problem.” It’s Sunday night at 2:00 in the morning or Monday morning at 2:00, I don’t wanna deal with it anymore. We’re rolling back. We’ve made the decision. Because when you start thinking about your who, what, when, where, why and how you’re going to ask yourself questions like how do we know we’re successful, at what point do we say we’re we’re putting the stick in the sand and we’re saying, “This is the line.” Where’s your red line and how do when you’re at it and who’s watching it? Because as technologists, I have a blog post out there, maybe Mike and my team will find it, but I basically talk about don’t forget to fly the plane. And I talk a lot about aviation disasters as well. As technologists and engineering type people we love to get into the nitty gritty and we might get so lost trying to fix a problem that the sun starts coming up on Monday morning and “Oh no, we’re up the creek.” So you have to have somebody watching the time and being willing to say “No, we’re rolling back.”

And with a migration upgrade rolling back as you change your CNAME over you say “OK, we didn’t hit the upgrade this weekend.” Instead of saying “Oh no we’re in trouble” and then feeling bad and sort of getting dejected, you say “We failed. Great. We learned what didn’t work. Now let’s go forward and learn from those mistakes and make it even better try next week.” Yes we had a downtime window. There’s a little opportunity costs or some sunk costs there that you might get worried about, “Oh no they’re going to hate me.” But your users I guarantee you will hate you a lot more if you kept going and going and going and you lost data.

“Do you have any specific steps, plans, tasks in Excel you’ve prepared?” Yes actually. I’m working on a document for that and we’ll probably have that checklist next week. So that’s one of the downloads that we have a slide already talking about. But again I can’t give you all of the steps but I’ll give you some of the steps that I like to use. So that’ll be in a download next week with a session next week on the doing but definitely have a checklist. Checklists are really important and we’ll get to that in that slide. So I’m gonna keep going because if I stay on one slide I’ll get lost but ask questions and let me interact. Again, the whole purpose of why I’m doing this and why I want to continue doing webinars for as long as you keep coming to them is to interact, to ask questions, to challenge me because I’m going to learn from you. So let’s just keep the dialogue going if you have questions.

So this is another good quote. “To remember where you come from is part of where you’re going.” So that quote should scare you. I bet you Ron, like Joseph from last week and judging from you talking about SQL 2000. I bet that quote scares you. You need to know where you are. You need to know what’s good in your current environment, what’s bad in your current environment. So even before you start with a migration plan, even before you pick your hardware or pick your cloud provider or pick your virtualization layer and spec out your servers, you need to know what’s good and what’s bad in your current environment.

So again, I’m not trying to talk about my company to try and sell my company. I’m talking about my company because it’s an example of how I do this all the time for customers. When we do an upgrade and you can do this yourself with tools like Glen Berry’s Diagnostic Scripts, Brent’s Split Script, you don’t have to call us. We don’t need to do your health check. You can do your own health check. You’re here. You figured out how to come here. Send me an e-mail. And in fact I’ll give you a download in an e-mail after this that links to our self-assessment PDF that you can just take on your own and not have to give me your email address for. But the point though is, do some assessments, look at your environment and understand what’s good and what’s bad and then look at your performance.

I have a blog post that talks about using the PAL tool, the performance analysis for logs tool, that’s a free Perfmon based tool that can do some analysis and you can use as a baseline. If you have Century 1 or Red Gate SQL Monitor or any of those third party tools that do monitoring. The point is to start doing some monitoring. Look and see what your resource utilization is. But don’t just look at your resource utilization, also look at your queries, look at your settings are you using more resources than you need to because one of your feet is on the gas and one of your feet is on the brake. So are you hurting yourself and is that making you use more resources than you need to.

So do this health assessment and every time we do an upgrade or migration I do a health assessment first. And sometimes the customer will say, “Well why would you do a health check for us? We’re getting rid of the server.” And I do the health check first because I want to see what you have and I want to see what’s different and what’s unique about your current environment. Because maybe you forgot that you have a default collation of something other than the normal default collation in your instance and you have four databases that have it this way and two that have it this way and you forgot that “Oh yeah I have this linked server that goes to this DB2-AS400 environment or something. And by doing your health assessment you get to analyze what you have and make sure that you’re thinking about all these things. So the health check really gets you in the mindset of sort of doing a, for lack of a better word, a tabletop exercise of thinking about when I do my upgrade what’s going to go wrong, what’s going to go right, where am I hurt today.

So a couple of things you can do is do a health check, do some performance analysis and don’t forget to look at the licenses your using. If you’re on 2000 or like Ron said if you’re on 2008 and you bought your licenses earlier with software assurance the licensing scheme has changed.

So do you need to be on Enterprise? We talked about that last week, I won’t rehash that. But SQL 2016, actually I will rehash, I’m about to. SQL 2016 SP1 gave us a lot of Enterprise features in standard. SQL Server 2014 and higher we can use 128 gigs RAM in standard. I’m in the middle of teaching an HADR class and some of my students are here from that class this week. You can do a failover cluster and then you can do a transactional replication for reading and you can have H.A. and you can have a read only replica through the replication without having to buy Enterprise Edition to use availability groups. You get failover cluster for H.A., you get replication for reading and you can do log shipping for D.R. and log shipping is perfectly respectable. When we do our webinar series on HADR I’ll talk about that. Log shipping deserves more respect. So do you need to be Enterprise Edition.

And don’t forget to look at your licenses and look at what kind of CPUs you can do and don’t just buy. “I don’t know what we’re using and I hope we don’t have a bunch of performance problems so let’s go with the biggest and beefiest box and get 482 cores just in case we need them.” That’s going to hurt you. So we want to try and be a little more spartan.

Now if we’re building on physicals we might want to get it closer to being right because on a physical we can’t just go and change CPUs if we realize that we got it wrong. On a virtual that has proper allocation and has enough headroom of more CPUs we can get it a little bit wrong and add CPUs. So I would start smaller CPU core wise and do some performance testing before you upgrade and watch and see what you need and we can add CPUs.

And then the cloud. If you’re building a VM in Amazon or in Azure you can get it wrong all day long and just reconfigure your server and have a brief blip and change your server from a small box to a gigantic box with minimal effort and minimal downtime. So we can get it a little bit wrong but we want to do this performance analysis to get it kind of right.

So another good quote that I like is, “If you don’t know where you’re going you’ll end up someplace else.” Isn’t that true? I mean and I’m preaching to the choir but how often do our companies do this to us? How often do our managers and our project managers and our vendors give us a marching order, and hopefully nobody from my team says, “All the time.” But give us a marching order and say, “We’re going this way, we don’t really know where we’re going but we’re gonna go this way.” And there’s no plan. There’s no goal. There’s no, “Here’s where we are now, here’s where we’re going. And this is the road to get there. This road can’t get there. This road can’t get there.”

Another saying out there is, “If you don’t know where you’re going you’ll end up there.” So again from the homework last week did you pick a SQL version? Again, I prefer SQL Server 2017 I think for most new environments that are calling me. 2019 is kind of around the corner but just based on my experience I think 2017 is a good place for most people to go. Even 2016 is an improvement over your SQL 2000 and 2008. If your vendors are a little slower and they’re getting nervous about 2017 you have more time before end of life for those two.

Are you going to new hardware? Hopefully you’re not doing an in place upgrade. Is it time to move to a virtual environment? Is it time to look at the cloud? And we’ll do a little sidebar chat at the end here if there’s time for the cloud to talk about that. But ask these questions today. Don’t ask these questions one week before go live, or even one month before go. Ask these questions now and figure out what are we about to do and how are we going to get there and why. And again go back to the questions I asked earlier; who, what, when, why, where and how. And all those questions, as you start asking them over and over again, and you could even ask underneath so like why, why, why, why? Like ask the why and then ask the why about the why and keep asking whys until you get to the discrete tasks and discreet questions and worries you should have and these become your checklist. And again we’ll give you a simple checklist next week. But that checklist is not going to be end all be all. It’s not

going to be end all be all it’s, not going to be your checklist; it’s my checklist and it’s generic, and you have to fill in the details and the gaps, because your environment’s different, and that’s why you have to spend this time not doing work, not troubleshooting problems, but asking these questions and thinking about your upgrade.

I like this quote too, “The most exciting phrase to hear in science, the one that heralds discoveries is not eureka, but now that’s funny.” That’s from Isaac Asimov. We got to test; the point here is we have to test. We shouldn’t … It’s exciting but not really a good exciting, necessarily, when it comes to a SQL server upgrade. That’s a cool discovery, like when you’re playing with moldy bread and you discover antibiotics, it’s cool when you discover how radiation can be used for medicinal purposes and discovery purposes; it’s not a great thing to say at T-1 second, before go live in the middle of an upgrade like “Oo, I forgot about compatibility level. Oops, I forgot about the..

Read Full Article
  • Show original
  • .
  • Share
  • .
  • Favorite
  • .
  • Email
  • .
  • Add Tags 

Separate tags by commas
To access this feature, please upgrade your account.
Start your free month
Free Preview