Discontinuous data appears
I thought it some problem on my side, bu may be i am not the only one
am I understand correctly - you received a lot of gaps?
or some metrics became disappeared at all?
Are only mysql graphs affected?
Just MySQL has problems with Linux normal
Can you check what you have on graph/dashboard/db/prometheus dashboard now and Before update
And additionally provide first lines of Exporter :42002/metrics-hr
for InnoDB Pare Reorgs - it’s not gap - it’s 0 as value, because you don’t have this type of activity in this moment.
Also what is your Interval value - Auto ?
probably we have the same problems. https://www.percona.com/forums/questions-discussions/percona-monitoring-and-management/48995-pmm-is-losing-instances
Also, Prometheus has changed their memory usage. So, for now, it’s possible that you have some limits for memory.
Please, reinstall pmm-server with manually set Metrics Memory limit https://www.percona.com/doc/percona-…upgrade-server
docker run -d
If you already used some METRICS_MEMORY try to increase it
@roma.novikov It’s normal now
i got same pb,could you share what you did to fix this please?
thanks in advance.
Hi romber ,
can you try to stop and remove current pmm-server instance and start a new one with tunned METRICS_MEMORY variable?
docker run -d
thanks for you kindly help.
when i remove pmm-server and reinstall it ,
do i need to repair and config all of pmm-client?
during pmm-client adding data about clients become stored inside consul for exporters and inside mysql for QAN.
mysql and consul databases stored inside pmm-data container.
so if they are ok (not broken), deletion of pmm-server container cannot affect on pmm-client configuration.
thanks for your help,it really helped me.