As we were building SQL ConstantCare®, I wanted it to be mentoring, not monitoring. I wanted to take daily data samples, trend them over time, and give you big-picture advice on what tasks you should focus on to make your SQL Servers faster and more reliable.
I envisioned it having two parts:
- Our robots would give you daily tactical advice
- I’d check in weekly to give you bigger-picture strategic advice
Over the last year, I learned a lot!
Our robots are keeping you pretty busy. The basic care and feeding of a production SQL Server is more challenging than most folks expect. (For example, over the last week, 391 servers have databases with no recent corruption checks, and 168 servers have databases without a full backup.)
Our robots are getting smarter, too. Over the last couple of months, we’ve been gradually rolling out query plan advice. If you’re sending in query plans, then each Monday, we tell you which queries to focus on tuning, and what changes you need to make.
For example, if we detect that your most resource-intensive queries are suffering from implicit conversion, we tell you, and we include the query plans attached to the email so you can see exactly which ones to fix.
That’s still in early access because we still have work to do on that front. For example, early access user feedback noted that:
- We need to add the database name in the query plan file name to make it easier to know which queries can be ignored (like from 3rd party apps, or known in-house tools)
- We need to filter out commonly known queries & databases (for example, you can’t really tune Ola Hallengren’s maintenance scripts, or by popular monitoring tools)
- We need to filter out really low-impact queries (because some servers are just flat out bored, and there’s no return-on-investment in tuning them)
- We need to let you quickly mute specific queries (as in, yes, I know that query is terrible, and we have no plans to fix it because the person who wrote it left, and we’re replacing that part of the app)
You don’t necessarily need the video training bundled in. A lot of folks know what a particular warning means, and for warnings they haven’t seen before, they have plenty of resources available in the more-info links in each email anyway. We could save them money by not bundling the video training if folks don’t need it.
You don’t necessarily need the active mentoring, either. In a lot of shops, even the most basic robot alerts are keeping your hands so full that you don’t need to hear me check in. In many others, the servers just aren’t having bad performance problems, so there’s no sense in spending time diagnosing them – users just bought SQL ConstantCare® to keep an eye on the server’s overall health.
So we’re relaunching it: simpler, with an even lower price.
That’s one easy-to-digest price that covers daily monitoring of all of your SQL Servers, regardless of their location. As always, it’s priced by email address that we send the alerts to – and we’ve found that most shops are just sending the alerts to a single distribution list email, and that keeps their costs down.
As existing subscribers approach their renewal date, their subscription renewal pricing will automatically be updated to the new lower price in the my-account page. (Europeans, fret not: we’re working on a GDPR-friendly release, and I’ll write about that in next week’s post.)