Replies: 1 comment 1 reply
-
You need a permanent storage for store metadata, only the parquet files in s3, but we have metadata in local disk as default. it would help if you mounted a volume to https://openobserve.ai/docs/storage/ |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I'm deploying Openobserve in my kubernetes cluster and I think I successfully set up an S3 storage for it. At least, once the pod is alive I can see the following structure created in S3 bucket:
files / myfolder / logs / logs / 2023 / 12 / 26 / 23
and many other folders. Every folder has a non-empty parquet file in it.But I tried restarting the pod and it doesn't seem to load old logs from s3. I can't find references in the docs suggesting if it's an expected behavior or not.
Beta Was this translation helpful? Give feedback.
All reactions