The current DB is a cloud server.
The pmm and mysql servers are separate.
However, the CPU part provided by pmm only shows the contents related to the current node (PMM server).
I want to see DB CPU usage.
The PMM server has plenty of CPU and memory, so this is unnecessary information.
What should I do to see the DB server CPU utilization (DB - Cloud, PMM access through bitvise)?
In the monitoring provided by the cloud server, the cpu usage rate exceeded 100%, but pmm cannot confirm this.
Which procedure did you use to add the server to PMM? If you add your server using the GUI, you won’t be able to view the Linux metrics in the PMM server; however, if you use the PMM agent, you will be able to view the OS metrics dashboard.
I have installed pmm with docker and am using pmm-agent.
However, the pmm server and the DB server are separate.
Currently, only the OS for the pmm server can be checked,
I want to check the OS or CPU usage for the DB server.
What Cloud do you use for your DB?
Can you install the pmm-client on the cloud machine you using for DB following instructions from Client - Percona Monitoring and Management as this is the best way to get the OS metrics.
but, as @Naresh9999 mentioned, you should be able to see the OS data on your version of Node Summary
In Node selector you should have the DB host name
Currently, naver cloud mysql is being used, and it cannot be installed separately in the cloud.
Currently, mysql_exporter is receiving the collection of mysql.
Is there any way to import without installing an agent in the DB server?
If you’re using “agentless” monitoring we only have the ability to gather metrics based on what can be obtained from mysql directly (i.e. through SQL queries).
The only way around this I know of would be if you were using Percona Server for MySQL (as opposed to community version of MySQL) where we’ve written a ProcFS plugin that will actually scrape the /proc directories of the DB server and ingest the metrics into a table in mysql…you can then configure an external exporter (from your PMM server) to query these metrics. Our founder wrote a blog detailing how it’s done here but there is a small hit to your DB server doing periodic ingestion of filesystem data and writing to disk.
In the end, it will be difficult without installing the agent directly on the DB server.