Comments on: How to Find Query Slowdowns Using Percona Monitoring and Management https://www.percona.com/blog/how-to-find-query-slowdowns-using-percona-monitoring-and-management/ Tue, 03 Nov 2020 08:13:32 +0000 hourly 1 https://wordpress.org/?v=6.5.2 By: Kay Agahd https://www.percona.com/blog/how-to-find-query-slowdowns-using-percona-monitoring-and-management/#comment-10972821 Tue, 03 Nov 2020 08:13:32 +0000 https://www.percona.com/blog/?p=71807#comment-10972821 Thanks Peter for your explanation!

]]> By: Peter Zaitsev https://www.percona.com/blog/how-to-find-query-slowdowns-using-percona-monitoring-and-management/#comment-10972817 Mon, 02 Nov 2020 18:55:48 +0000 https://www.percona.com/blog/?p=71807#comment-10972817 Kay,

It is kind of tricky. The “bad query” which generates a lot of IO and overload disk subsystem is itself likely to additionaly suffer from such added IO latency as well as have “good queries” to suffer.

In many cases it is good idea to think about what is normal and what is expected – if disk IO latency is significantly elevated from normal it is likely caused by queries causing disk IO which you can find and analyze.

]]> By: Kay Agahd https://www.percona.com/blog/how-to-find-query-slowdowns-using-percona-monitoring-and-management/#comment-10972767 Sat, 10 Oct 2020 20:51:19 +0000 https://www.percona.com/blog/?p=71807#comment-10972767 Daniel, interesting article, thanks! One question though: Could you please elaborate why you concluded that’s it’s not an issue with the query itself only because I/O load went up which you describe as as a “Disk problem”? I mean, if a query is slow, the reason is often that it had to access the disk, hence I/O load will go up. The question hence is rather, why it had to access the disk. Did the size of the index increased so that it can’t fit in RAM anymore? Did the size of the data increased? Did the access pattern changed? Etc. pp.

]]>