First of all, I am sorry that my English is not smooth.
I manage MySQL 5.7 version on public cloud with PMM.
Because of cloud envrionment, I install exporter on PMM Server, not on a client Server.
And I only activated mysqld_exporter except linux_exporter.
By the way, mysqld_exporter often doesn’t seem to work.
I attach two screenshots containing the corresponding phenomenon…
Please Tell me what I have to do.
What’s in the log file at /var/log/pmm-metrics* ?
I upload it again with the screenshot of log as the most recently broken point.
After when the collector scrapes “perf_schema.file_events” and “info_schema.processlist” Error occured, pinging error is continuously occurring.
The error log appears that the exporter is having issues connecting to the database. However the database metrics from MySQL Overview dashboard show that the server isn’t that heavily loaded (threads_running ~ 4). The error message reads that the exporter is timing out connecting to MySQL.
There is Prometheus Exporters dashboard - can you look at that and see if scrapes from mysqld exporter are taking a long time for this host?
One option is to set the scrapes to occur less frequently. look at METRICS_RESOLUTION and set this to perhaps 5s
[URL]Percona Monitoring and Management
I ran the pmm docker with the -e METRICS_RESOLUTION=5s option.
I upload screenshot of Prometheus Exporters dashboards that contain the time when the error occurred.