Since last HA update, integration raises error messages

[bdauvissat]

Environment
Home-Assistant version: 2024.12.1
Elasticsearch version: 8.15.3

Describe the bug
Since my last update, the Elasticsearch integration raises errors.

I tried to remove it, it failed.

I deactivated it and then removed it, it worked.

When I try to install it again, I have the errors in the attachment

home-assistant_2024-12-09T13-59-30.177Z.log

[strawgate]

This appears to be this issue: #403 (comment)

You could upgrade to the 2.0 pre-release version by enabling the "beta" list of versions when upgrading. The 2.0 version changes the document format which will break your visualizations and dashboards.

In the meantime, I'm trying to see if we can fix the v1.0 version so you won't have to change your content.

[bdauvissat]

I tried with 2.0 RC3
Nothing appears in my cluster.
I tested auth with API key and user/password, same issue.
Please find attached logs
home-assistant_elasticsearch_2024-12-09T15-14-56.021Z.log

[strawgate]

Okay, I see two issues in your log. One is related to the area definition and the other is a routing issue.

For the routing issue:
The error:

2024-12-09 16:14:30.793 ERROR (MainThread) [custom_components.elasticsearch-httpshomeassistanteseucentral1awsfounditno] failed to create document {'_index': 'metrics-homeassistant.sensor-default', '_id': None, 'status': 400, 'error': {'type': 'illegal_argument_exception', 'reason': "Error extracting routing: source didn't contain any routing fields"}}

Only makes sense if the underlying datastream is not a TSDB datastream -- can you check if the metrics-homeassistant index template exists and if the metrics-homeassistant.sensor-default datastream exists or if its a plain index? Could you try deleting the metrics-homeassistant.* datastreams and reload the integration? Are there any errors in your Elasticsearch log?

For the area issue:
Can you share the configuration of the area assigned to your freebox entities?

[strawgate]

Rolling over your datastreams should fix the issue with routing and there is now a PR up for the area issue

[bdauvissat]

After rolling over my datastreams, I don't see routing issues anymore.
Index template exists, so as the datastream (it's not a plain index).
No errors in the cluster logs.
I don't know where I can find the area information you're asking for.

[strawgate]

That's okay, I reproduced the area issue and have an open PR to resolve it. Will comment here once there's a new version to try