PMM monitor image is discontinuous,and how to modify the value rating
Not quite sure I understand the question. Is this in relation to AMI or OVF images? If you can provide a bit more detail we can probably get you pointed in the right direction!
Hi @lee it looks like your PMM Server is unable to keep up with the rate of scraping you’re asking it to do. By default in PMM1 there are generally three resolutions per exporter:
- lr low resolution 60s
- mr medium resolution 5s
- hr high resolution 1s
https://www.percona.com/doc/percona-monitoring-and-management/glossary.option.html#passing-options-to-pmm-server-for-an-already-deployed-docker-instance
There’s a Prometheus dashboard in the 1.x release series that will show you if your system is overwhelmed. It is usually the disks can’t keep up, but you might also benefit from more CPUs and additional RAM.
Also you are running quite an old version of PMM Server at 1.6. Please consider upgrading to PMM2, or at the very least to latest in the 1.x release. Notably we improved Prometheus performance around 1.12
In PMM2 you can see the performance of the exporters:
https://pmmdemo.percona.com/graph/d/prometheus-status/prometheus-exporter-status
And of Prometheus itself:
https://pmmdemo.percona.com/graph/d/prometheus/prometheus
thanks for your ans. i also try used pmm2 or latest in the 1.x release.but the OS is linux 5.8 and 6.x. and mysql release is 5.1,can u support some suggestion?
MySQL 5.1 is not something we support with PMM (as in test PMM with it) though some graphs may still work as in your case.
The status you see does corresponds to Prometheus overloaded - note you have some “Last Scrape” being 14min ago + this points to Prometheus reducing scrape frequency due to overload.