MySQL RDS graphs are empty & hosts servers have disappeared in grafana

It took awhile to get PMM connected to the pmm-server (there was a gap in the documentation for RDS). I installed the pmm client software on the ec2 instance that is hosting the pmm-server. I did not put it in a container. I ran pmm-admin config for the local server. Ran pmm-admin add mysql:metrics for 3 RDS instances. And suddenly there were graphs with data. Great data. Pretty graphs. Useful.

About 24 hours later my graphs disappeared. I have followed a couple of threads in here where people lost their graphs. Here’s what I’ve done:

  1. https://myserver-ip/prometheus/targets Shows they are ALL down (including http://localhost:9100/metrics )

  2. https://myserver-ip/42002/metrics-mr (plus lr & hr) all show data. If I hit refresh the mysql_exporter_hr_scrapes_total number increases

  3. Main page where the v1.1.3 is & hit the blue button which did an update (today May 19, 2017) - nothing changed.

  4. Drop down for the MySQL Host has None & a large blank box. Manually add the aws host ID into the box and a couple of graphs (MySQL Connections but No MySQL Active Threads) have old data.

  5. One of the forum posts talked about changing the scrape_timeout: in prometheus.yml in the docker container. My docker container does not have a prometheus.yml file.


Hi Erino,

Is http://localhost:9090/prometheus/metrics target also down?
in this case please synchronize time (NTP or chrony)

Hello! This was fixed by adding resources to the ec2 instance! I found the requirements of PMM that Peter had written (google PMM & blog is how I found it). Thanks so much for your help. I LOVE PMM!