I’m contacting you today because I’m experiencing problems on one of my payara servers (a test server). Sometimes, payara-server doesn’t respond to webservice requests. At first I thought the problem was with our reverse-proxy, but in consultation with our IT department, we ruled out this possibility.
I set up a grafana dashboard to monitor webservice calls. Every 5 minutes, I call a test service, which simply returns “OK” so that I can retrieve the http response code. This dashboard has enabled me to see that the time during which payara is unavailable is between 1h05 and 1h10. If we remove the bias linked to the fact that I only monitor every 5 minutes, the unavailability would systematically be 1 hour. This made me think of a blacklisting mechanism.
- host system : debian 10
- docker CE : 20.10.18
- container system : ubuntu 20.04.4 LTS
- Payara : 5.2022.2
- JDK : Zulu17.32+13-CA (build 17.0.2+8-LTS)
My question: does Payara have a blacklist mechanism if a lot of WS calls fail?
Please note that during webservice unavailability, the admin console is still accessible.
Thanks for your help
Here are a few screenshots to illustrate the problem.
Note: On the second one, around 4 p.m., the data gap corresponds to a container shutdown to rebuild a new docker image.