Timeout when listing indices in the dashboard

Versions (relevant - OpenSearch/Dashboard/Server OS/Browser):
OpenSearch version 2.13.0
Dashboard version 2.9.0

Describe the issue:
Whenever I try to load the indices through the index management panel, I get a 30s timeout error.
Does anyone know what cluster setting should i alter to increase or disable that timeout.
Thanks

Configuration:

Relevant Logs or Screenshots:

1 Like

Hey @YassineLazaar

It should l show you index sets. Was it like this before or did this just happen?

Edit:

These should match versions

Not sure, probably it was always like this.
We had matching versions be we had to rollback the dashboard because of a link sharing issue.

You may want to enable logging in your opensearch_dashboards.yml file.

logging.dest: /var/log/opensearch-dashboards/opensearch-dashboards.log

That may help you find more info about what’s happening. Additionally, I think Dashboards sometimes will send more information to your browsers developer console.

30s is a long time for it to load that information. I would guess that it’s more likely there’s something wrong with OpenSearch itself, rather than Dashboards.

Edit: You could also run a simple API call in Dev Tools to see if that is faster. If it works instantly than that would indicate a Dashboards issue. If it’s slow too, that would indicate an OpenSearch issue.

GET _cat/indices
GET _cat/indices

This has also failed either with the dev tools or curling the opensearch directly.

The cluster itself looks healthy

What did _cat/indices return when you used curl?

Does the related _cat/shards work with curl?

I scaled the cluster to 5 instances with slightly reduced memory. It reached a state where the indices and shards are returned with curl but we still get a timeout when listing with the dashboard.

1 Like