I am having 30+ kong container running which are connected to same Postgres database. I wanted to know, Do all my 30 kong Instances check for Active Health Check when applied in upstream entity? These containers keeps on autoscale whenever there is a load so this 30 number is not fixed. If Yes, then Is there a way to restrict it to single kong Container.
Each Kong container will health-check the upstream and decide independently if it should be routing traffic to a particular target of an upstream.
This helps avoids a failure scenario where requests won’t be routed to an unreachable target during network partitioning or congestion.
Thanks Harry, But I think Prometheus Plugin is calculating these requests also, which ultimately giving us really high number of requests in comparison to the actual requests. So Is there anyway to restrict this count ? And in Proxy Logs, I can just see the health check requests which fails, the requests which get passed, those are not present in the logs.
Please don’t mix up Prometheus plugin and Active health-checking. Those are two separate components and don’t share any data.
I’m not sure what your question is here.
Sorry for late reply. so my question is how can I lower the count of this active health check, as I am having large number of kong instances connected to same db?
Have you considered lowering the interval for these health checks? That would reduce the number of health-checks that are performed but that also means that it will take longer for Kong to detect a change in health of an upstream if it goes down.