Updated First Responder Kit: sp_Blitz markdown output, sp_BlitzIndex statistics checks

In addition to lots of bug fixes and tweaks, my favorite option is:

sp_Blitz @OutputType = 'markdown'
sp_Blitz @OutputType = ‘markdown’

sp_Blitz @OutputType = ‘markdown’, @CheckServerInfo = 1, @CheckUserDatabaseObjects = 1

See, recently I’ve spent a lot of time looking at Stack questions going, “Man, if I just had the output of sp_Blitz, I bet I could answer that question in five minutes.” But sp_Blitz’s output doesn’t lend itself well to copy/pasting.

StackOverflow lets you format your questions and answers with a flavor of Markdown, a cool way of editing text files. So I added @OutputType = ‘markdown’ with a bullet-list-tastic format. (Sadly, Stack’s flavor of Markdown doesn’t support tables, which would have made things easier.)

I don’t include all of the fields from sp_Blitz’s results – for example, there’s no URLs or query plans, because those things don’t make as much sense here.

In other news:

sp_Blitz v53.3:

  • New check for failover cluster health (Matt Tucker) – warning if you don’t have failover cluster nodes available in sys.dm_os_cluster_nodes.
  • New check for endpoints owned by user accounts (HBollah)
  • New check for wait stats cleared since startup (Brent Ozar) – useful in combination with @CheckServerInfo = 1, which includes any wait stats that are bottlenecks.
  • Improved Amazon RDS and SQL Server 2005 compatibility (Brent Ozar)

sp_BlitzFirst v25.2:

  • Split what’s running now code into new sp_BlitzWho (Ryan Howard) – several folks said they found this section useful on its own. We’ll add more into sp_BlitzWho down the road.

sp_BlitzIndex v4.2:

  • Added statistics checks (Erik Darling) – warning about outdated stats and stats with low sample rates.
  • Added database name to multi-db result sets (Brent Ozar) – notable because it’s a breaking change if you’d built anything on top of @GetAllDatabases = 1.

Go get the goods now, and if you’d like to contribute code or file issues, head over to the Github repo.

Previous Post
Questions You Should Be Asking About Your Backups
Next Post
[Video] Office Hours 2016/10/12 (With Transcriptions)

4 Comments. Leave new

  • I’ve been thinking … is there an easy way to keep this up to date on my test server? I know there is a github repository, but don’t know of a way to import directly without copy/pasting.

    Is there an easy way I don’t know of that would let me essentially “check for updates” from within SSMS?

    • Andrew – we actually did this before as a fun experiment. I built an sp_BlitzUpdate as a proof of concept and showed it at the PASS Summit. It used a linked server to go fetch an updated definition of sp_Blitz, and then alter it.

      The security concerns are pretty obvious: you don’t really want to automatically get code from a stranger’s server and run it live on your production SQL Server. That’s way too much risk.

  • Makes sense. I was just curious about the possibility of it. I’m setting up our first utility DB at work, and putting a lot of my own spin off/logging procs as well as some base Hallengren scripts, and the First Responder kit so it is all easily redeployable quickly when we spin up a new box. Would be nice to always have latest versions because they will come from a variety of sources. Obviously I had thought about going to isolated test environment first and then having SSIS push out to production after testing. I might still try it with a SSIS package with a fetch/unzip script object just to see if I can do it. For gits and shiggles.

    Looking forward to seeing you at PASS this year, I’ll come up and say hi!

  • Great to see Statistics getting a mention in sp_BlitzIndex. I have never understood why the SQL Monitoring tools don’t make more use of this. Good job Erik!


Leave a Reply

Your email address will not be published. Required fields are marked *

Fill out this field
Fill out this field
Please enter a valid email address.