Hide Forgot
Description of problem: there is unnamed query which runs a lot and hit the performance. this query found when running CFME with RHV. LOG: duration: 3059.466 ms execute <unnamed>: SELECT history_id, history_datetime, current_user_id, current_user_name, cast(user_logged_in_to_guest as int), vm_id, seconds_in_status, cpu_usage_percent, memory_usage_percent, user_cpu_usage_percent, system_cpu_usage_percent, vm_ip, vm_client_ip, currently_running_on_host, vm_configuration_version, current_host_configuration_version FROM vm_samples_history WHERE vm_status = 1 AND history_datetime >= '2016-12-04 22:00:00.000000+0000' AND history_datetime < '2016-12-04 23:00:00.000000+0000' ORDER BY history_datetime, current_user_name, vm_id Version-Release number of selected component (if applicable): 4.0.6.1 How reproducible: 100% Steps to Reproduce: 1. scale out lab with 27 hosts and 2.2K vms 2. 3. Actual results: non-optimized query Expected results: optimized query. Additional info:
This query rum every hour. Unless for some reason there was an issue with aggregation and it is now running to catch up to the hour before last.
(In reply to Shirly Radco from comment #1) > This query rum every hour. Unless for some reason there was an issue with > aggregation and it is now running to catch up to the hour before last. any chance it caused by https://bugzilla.redhat.com/show_bug.cgi?id=1402471 ?!
For all we know, this is NOTABUG. Closing as such. Eldad - please re-open properly, with all information.
(In reply to Yaniv Kaul from comment #3) > For all we know, this is NOTABUG. Closing as such. > Eldad - please re-open properly, with all information. This bug caused by missing index's, somehow we can't find them on the relevant table "vm_samples_history" fixing it by re-running the crate_tables.sql script. im not sure, but I suspect that something was failed on the upgrade process. we'll update \ submit a new bug for it.