You need to speed up a SQL Server, but you’re not sure where to begin. Is it a database problem? A query issue? Not enough hardware? Goofy SQL Server settings? Good news – I do this for a living, and I’ll teach you how in 3 days of learning and fun. Join me, Brent Ozar, as I explain how to assess your SQL Server’s bottlenecks and figure out where to focus your tuning efforts.
You’ll work along with me in a VM – each student gets their own SQL Server 2019 lab VM in the cloud with the Stack Overflow database and running workloads. Each day is structured like this:
- 9AM-Noon – lectures, where we explain the concepts you’ll be working on
- Noon-2PM – lab and lunch. It’s up to you whether you work on the lab for an hour first, then go to lunch, or vice versa, or spend more time on the lab. Or have martinis for lunch. Whatever.
- 2PM-3PM – you watch Brent work through the lab you just did, explaining my techniques as I go.
- 3PM-4PM – afternoon lecture
- 4PM-5PM – lab homework (and you can also work on this later if you like – your VM will stay on overnight.)
This mix of lectures and labs helps reinforce the topics you’re learning and makes sure that you’ll be able to tackle the same kinds of projects in your own environment. (Many students do just that: after hours, they switch into their own environments and use the same diagnostic queries to make immediate improvements in their apps.)
You can choose to skip the lab VM – in the Mastering classes, I alternate between lectures and hands-on lab assignments. If you’re willing to build your own lab VM, you can save a ton of money. Just make sure you’ve got access to a SQL Server 2016 or newer with 4-8 CPU cores, 32-64GB RAM, and at least 600GB of free SSD space. Here’s how to set up your own lab workstation.
Labs & lectures include:
- How to use sp_BlitzFirst and the Power BI Dashboard to identify your server’s top bottlenecks
- How to fix common wait types like CXPACKET, PAGEIOLATCH, LCK%, SOS_SCHEDULER_YIELD
- How to identify and mitigate poison wait types like RESOURCE_SEMAPHORE and THREADPOOL
- How to prove that your changes made things better
- How to decide whether configuration changes, query/index tuning, or hardware/VM changes will give you the biggest bang for the buck
This is an online class – I host it in GoToWebinar, live with me on webcam, taking questions in Slack as we go through the material together. Audio can come through either your computer audio, or by dialing into a US phone number – headset recommended either way.
You can watch again later with Instant Replay – for one year from the date of your course, you can watch a recorded version via online streaming. Brush up on stuff you missed or revisit your favorite topics. It’s the best of both worlds: live training with me, plus reminders. (Or, you can just buy the Instant Replay access – save money, and start watching your recordings immediately, streaming them in your browser for a year.)
Prerequisites – this course won’t explain anything about index or query tuning, but we WILL be telling you things like, “For this particular wait, you’ll need to look for these types of queries, and tune them by doing X.” We won’t explain those concepts here – we’re only going to be looking out across the entire server, and figuring out what tasks we need to perform. In some cases, you’ll be able to fix the problem just with hardware or SQL Server settings, but be aware that that’s not usually the case. Therefore, for the best possible experience, we’d recommend taking both Mastering Index Tuning and Mastering Query Tuning before tackling this class. After you enroll, do the full prerequisites including chat setup.
Class dates & times:
- May 22-24 (Fri/Sat/Sun), 2020 online 10:00-4:00 Eastern (15:00-21:00 UTC) (click to see in your own time zone)
- June 9-11 (Tues/Weds/Thurs), 2020 online 9:00-5:00 Eastern (14:00-22:00 UTC) (click to see in your own time zone)
- Sept 9-11 (Weds/Thurs/Fri), 2020 online 9:00-5:00 Eastern (14:00-22:00 UTC) (click to see in your own time zone)