Getting CrashLoopBackOff for minio-0 pod

[devopsn3rd]

1.- Description
When running the installation script, the operator runs and the minio-0 pod won't start. It fails with "CrashLoopBackOff".

2.- Environment
GCP with Ubuntu LTS 18.04. 4 CPUs and 8 GB memory.

3.- Some troubleshooting

i.- kubectl describe pod minio-0

Events:
Type Reason Age From Message

---

Warning FailedScheduling 5m1s default-scheduler persistentvolumeclaim "minio-pvc" not found
Normal Scheduled 4m51s default-scheduler Successfully assigned spinnaker/minio-0 to minnak
er
Normal Pulled 4m11s kubelet Successfully pulled image "minio/minio" in 38.871
741781s
Normal Pulled 3m49s kubelet Successfully pulled image "minio/minio" in 398.24
3525ms
Normal Pulled 3m12s kubelet Successfully pulled image "minio/minio" in 362.38
4058ms
Normal Pulled 2m47s kubelet Successfully pulled image "minio/minio" in 357.34
6151ms
Normal Created 2m47s (x4 over 3m51s) kubelet Created container minio
Normal Started 2m47s (x4 over 3m51s) kubelet Started container minio
Warning BackOff 2m19s (x7 over 3m28s) kubelet Back-off restarting failed container
Normal Pulling 2m5s (x5 over 4m50s) kubelet Pulling image "minio/minio"
Normal Pulled 2m5s kubelet Successfully pulled image "minio/minio" in 369.44
3169ms

ii.- kubectl logs minio-0
/bin/sh: /usr/bin/minio: No such file or directory

iii.- I tried the workaround described in issue #99, but it didn't work

kubectl apply -f $BASE_DIR/infrastructure/minio.yml

Note: I replaced for my own base directory.
Then:

kubectl delete pod minio-0

And it fails with the same error.

Regards,
Luis.

[mattbarlow-sg]

What worked for me was editing spinsvc/infrastructure/minio.yml, changing the path from /usr/bin/minio to /opt/bin/minio and then running ./deploy.sh, then deleting the old minio-0 pod.

[hsmade]

This happens because the minio/minio image is not pinned, so it defaults to :latest :(

[tim1670]

I also encountered the same problem

[ibuamod-tc]

I also encountered the same problem

[laurie-kepford]

So did I. The fix of changing the path from /usr/bin to /opt/bin worked as far as getting it to deploy. But it still does not work. When I try to go to the URL I get a 404 error. And the logs for minio give me a warning about it running on a dynamic port. And it also says: WARNING: MINIO_ACCESS_KEY and MINIO_SECRET_KEY are deprecated

[ravikiranr26]

What worked for me was editing spinsvc/infrastructure/minio.yml, changing the path from /usr/bin/minio to /opt/bin/minio and then running ./deploy.sh, then deleting the old minio-0 pod.

Tried the same but getting as follows from log of Pod, and when visiting the endpoint gives 404 page not found

root@minnaker:~/minnaker-0.1.3/spinsvc# kubectl logs minio-0 -n spinnaker
WARNING: MINIO_ACCESS_KEY and MINIO_SECRET_KEY are deprecated.
Please use MINIO_ROOT_USER and MINIO_ROOT_PASSWORD
API: http://10.42.0.17:9000 http://127.0.0.1:9000

Console: http://10.42.0.17:38189 http://127.0.0.1:38189

Documentation: https://docs.min.io

WARNING: Console endpoint is listening on a dynamic port (38189), please use --console-address ":PORT" to choose a static port