Open danielkurniadi opened 5 months ago
cc @lmenezes if you know or can point out quickly to any pointers. We have used this in production and now its blocking our rollout.
Here is what our workload looks like in Kubernetes. We have cerebro and we deployed elastic search for high volume data ingestion, more or less following this approach:
❯ kgd
NAME READY UP-TO-DATE AVAILABLE AGE
cerebro 0/2 2 0 12d
es-logging-exporter 1/1 1 1 427d
kibana-logging-kb 1/1 1 1 427d
logging-curator 1/1 1 1 427d
❯ k get sts
NAME READY AGE
elasticsearch-logging-es-data-optimised 2/2 249d
elasticsearch-logging-es-masters-optimised 3/3 249d
Description
We observed that the cerebro pod keep on crashing in the tm-gcp-emea-build-infra in namespace: logging during the start and there were issue related to one of the Slick java plugins. Checking in if this is a known issue that you've seen or is familiar with where you could point out the solution.
More logs in the gist here: https://gist.github.com/danielkurniadi/a3eb6cebdf64363ffbb9980b17714143
Configurations
The following is the configuration we added to
cerebro
.Can you help look and help provide more guidance on where to start looking on this? Some more information on what the issue really is can be really helpful in case you have observed or known this issue before.
Thanks 🙏