@cubejs/cubestore unable to start container in openshift

I have defined all the objects required in openshift starting with depolyment config pulling the image from docker hub cubejs/cubestore:latest. When starting a (pod)Container runs into CrashBackLoop with the below error.

thread ‘main’ panicked at ‘called Result::unwrap() on an Err value: ParseIntError { kind: InvalidDigit }’, cubestore/src/config/mod.rs:458:55
stack backtrace:
0: rust_begin_unwind
at /rustc/a143517d44cac50b20cbd3a0b579addab40dd399/library/std/src/panicking.rs:493:5
1: core::panicking::panic_fmt
at /rustc/a143517d44cac50b20cbd3a0b579addab40dd399/library/core/src/panicking.rs:92:14
2: core::option::expect_none_failed
at /rustc/a143517d44cac50b20cbd3a0b579addab40dd399/library/core/src/option.rs:1300:5
3: cubestore::config::Config::default
4: cubestored::main
note: Some details are omitted, run with RUST_BACKTRACE=full for a verbose backtrace.

It seems you specified non-integer value for CUBESTORE_PORT.
We’ll make sure to improve error messages in the future version.

Please post values for your CubeStore config vars if changing CUBESTORE_PORT doesn’t help.

1 Like

The variable CUBESTORE_PORT is unused, i haven’t set it. But i see that issue is coming up when i have serivce object running for cubestore before starting the pod. If i first start the pod and then the services. It doesn’t give me that issue

Below is the service manifest file.


apiVersion: v1
kind: Service
     name: cubestore
      - name: cubestore
        port: 3030
        protocol: tcp
        targetPort: 3030
       deploymentConfig: cubestore

What would be the target port for cubestore is it 3030 or 3306?
And do i need to make any changes in service.yaml manifest file?

You want 3030 (the http interface).
3306 is the MySQL interface, not used in production, we keep it only for debugging purposes.

The line in the backtrace is probably outdated then :thinking:Next build of CubeStore will have better error messages for env var errors (change landed today). But if that’s not your case, I’m not sure what goes wrong. Knowing the exact CubeStore version might help.

Note that running multiple routers of CubeStore concurrently that point to the same data can result in races.