# Percona toolkit find where load is coming from **URL:** \ **Category:** Percona Toolkit\ **Tags:** mysql, percona\ **Created:** [August 20, 2024, 6:32pm UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468 "2024-08-20T18:32:53Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![Idan\_Ahal](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/idan_ahal/32/13477_2.png) [@Idan\_Ahal](https://forums.percona.com/u/Idan_Ahal)\ **Post date:** [August 20, 2024, 6:32pm UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468/1 "2024-08-20T18:32:53Z")
I know the Percona toolkit has the option to analyze slow queries but can I use it to analyze and find what currently creates load on the MySQL? Sometimes we get into a situation where MySQL is on 100% CPU but we can’t understand why from the process list. Sometimes it’s just a lot of very fast queries. Can the Percona toolkit help me with that? Is it possible to understand what connection, user or query created what load on the MySql server? ---
**Author:** ![kedarpercona](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/kedarpercona/32/809_2.png) [@kedarpercona](https://forums.percona.com/u/kedarpercona)\ **Post date:** [August 21, 2024, 4:59am UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468/2 "2024-08-21T04:59:43Z")
PMM’s Query Analytics is best to analyze such spikes. ---
**Author:** ![Idan\_Ahal](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/idan_ahal/32/13477_2.png) [@Idan\_Ahal](https://forums.percona.com/u/Idan_Ahal)\ **Post date:** [August 21, 2024, 8:38am UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468/3 "2024-08-21T08:38:34Z")
Thanks for the replay! I read about PMM. Amazing tool. |But I’m looking for something smaller. Like running a tool that will output all the queries running for the past 5 minutes with time taken aggregated or something like that. Does `pt-query-digest` can do the job? I’ll just sell slow-queries bigger than 0, meaning all. Will it know how to aggregate the same query and know even if a very short query that runs a lot of times takes most of the CPU? ---
**Author:** ![kedarpercona](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/kedarpercona/32/809_2.png) [@kedarpercona](https://forums.percona.com/u/kedarpercona)\ **Post date:** [August 25, 2024, 5:53pm UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468/4 "2024-08-25T17:53:32Z")
Hi @Idan_Ahal, If you want to log all the queries consider adjusting the long query time. `set global long_query_time=0 # or 0.1` Collect the log for the desired period, ensure performance / disk usage are acceptable. You can rotate your slow log to easily parse them directly with the digest OR you can use the ` --since` and `--until` options to parse required information. The digest will not tell you what resource a query is using but you can get the stats for sure. Thanks, K ---
**Author:** ![Idan\_Ahal](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/idan_ahal/32/13477_2.png) [@Idan\_Ahal](https://forums.percona.com/u/Idan_Ahal)\ **Post date:** [August 26, 2024, 7:55am UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468/5 "2024-08-26T07:55:25Z")
> [@kedarpercona](#): > > The digest will not tell you what resource a query is using What do you mean by that? ---
**Author:** ![kedarpercona](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/kedarpercona/32/809_2.png) [@kedarpercona](https://forums.percona.com/u/kedarpercona)\ **Post date:** [August 26, 2024, 8:27am UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468/6 "2024-08-26T08:27:58Z")
> Will it know … query that runs a lot of times takes most of the CPU? No. It will just summarize the log for you. ---
**Author:** ![Idan\_Ahal](https://sea1.discourse-cdn.com/flex019/user_avatar/forums.percona.com/idan_ahal/32/13477_2.png) [@Idan\_Ahal](https://forums.percona.com/u/Idan_Ahal)\ **Post date:** [August 26, 2024, 11:51am UTC](https://forums.percona.com/t/percona-toolkit-find-where-load-is-coming-from/32468/7 "2024-08-26T11:51:51Z")
But if a query runs for 0.001 seconds, for millions of times than the total run of this query will be very big, which means it makes most of the load. Am I missing something?