We’re kinda like an emergency room for SQL Server: we specialize in a 3-day SQL Critical Care® where we work side by side with you, talking about your database pains, showing you the root cause, and then teaching you how to get permanent pain relief. That works really well, and we’ve kinda got it down to a science.
When we sign a contract with a client and pick a start date, we tell ’em it’s really important not to restart the SQL Server in the week leading up to the engagement. SQL Server keeps so much good stuff in memory – wait stats, file stats, index usage & recommendations – and all of that gets wiped out on an instance restart.
Nobody wants to restart their SQL Server instance.
But you know how it is.
Life finds a way.
Some well-meaning sysadmin applies a patch, or doesn’t know about the engagement, or folks just plain old run into an emergency and have to fail over.
So we thought, why not get new clients started with SQL ConstantCare® as soon as they sign a contract? We could just give them free access to start right away, sending in their server’s metrics every day.
This is so useful for so many reasons:
- Clients start getting advice faster – like warning them about backup issues, easy misconfigurations, and things they can fix without waiting for us
- We get better historical data – so useful in cases like a dramatically underpowered server that doesn’t have enough memory to keep plan cache contents around for more than a few hours
- We can spot things that only happen rarely – for example, if a server has a pattern where the wait stats look wildly different on Monday mornings, we can narrow that down and understand why
- And even after the engagement – we can keep an eye on a client’s server and know if they’re making progress on their homework, and whether they’re seeing permanent pain relief
My favorite example of how it’s been useful was a client that emailed in one Monday morning and said, “It just happened again! We had another performance-wrecking emergency on Sunday. Did SQL ConstantCare give you enough data to tell you what it was?”
Ordinarily, that’d be really hard to do as a consultant – I just don’t have the ability to time travel backwards and see what was going on. But since they’d been in the program already, I was able to just open up my Power BI dashboard, look at their data, and said, “Yep.”
It’s about to get even better: our beta customers with the latest ConstantCare.exe can now send in queries and query plans, too. We’re not giving automated query advice yet – that’s coming next – but at least we can capture the query plans so consulting customers can get better answers about which queries were causing the big slowdown on Sunday.