Fusion Auth Node getting removed.

lakshay

I am running Fusion Auth in my Kubernates Cluster. Using this fusionauth/fusionauth-app:1.28.0 image
I am using 512mb memory and 500m Cpu
The pods keep restarting after few minutes or hours randomly.
The only error or message getting logged is.

Node [1f98f15f-fd1e-4a34-b976-06be6ca89b2c] with address [http://10.2.206.72:9011] removed because it has not checked in for the last [102] seconds. Bye node.

I am using Aurora Mysql (AWS) database.

I am not able to understand why is this happening.

Is it something related to database connection (wait timeout or interactive timeout)
Some issue with Fusion Auth memory or cpu setting

If any one faced a similar issue please help.

Thanks

joshua

@lakshay,

Sorry to hear that you are having issues!

I am using Aurora Mysql (AWS) database.

To my knowledge, AuroraDB is not natively supported by FusionAuth (at least out of the box). See our system requirements. You can try adjusting your connection pool properties on the FusionAuth side, or updating settings on the Aurora side. My assumption is the HikairiConnection Pool (used by FusionAuth) and AuroraDB are not in agreement on what is connected and for how long. Tweaking settings on either side may alleviate your issue.

I am not very familiar with the K8s side of things, but we do have a repo devoted to user-supported submissions(here). You may find additional configuration assistance there if, in fact, it is not the database connection causing the issues.

I hope this helps!

Thanks,
Josh

dan

To follow on to what Joshua said, it might be helpful to review some of the troubleshooting docs: https://fusionauth.io/docs/v1/tech/troubleshooting/ and see if they help.