Fulltext Search container is unhealthy after AIO 13.2.0 Beta / Nextcloud 33.0.5 update

[vawaver]

Summary

After updating my Nextcloud AIO instance to AIO 13.2.0 Beta with Nextcloud 33.0.5, the nextcloud-aio-fulltextsearch container starts, but Docker/Portainer reports it as unhealthy.

The issue appeared immediately after the update. Before the update, the Fulltext Search container was working normally.

All other AIO containers are reported as healthy/running. Only the Fulltext Search container is unhealthy.

Environment

• Nextcloud AIO version: 13.2.0 Beta
• Nextcloud version: 33.0.5
• Host OS: Ubuntu 24.04 LTS
• Installation type: Nextcloud AIO
• Affected container: nextcloud-aio-fulltextsearch

Actual behavior

The nextcloud-aio-fulltextsearch container starts, but remains in an unhealthy state.

In Portainer, the container is shown as:

nextcloud-aio-fulltextsearch — unhealthy

The container is created and running, but the health status does not become healthy.

Expected behavior

The nextcloud-aio-fulltextsearch container should start normally and become healthy after the AIO update.

Steps to reproduce

1. Update Nextcloud AIO to 13.2.0 Beta.
2. Let AIO update/recreate the containers.
3. Start the AIO stack.
4. Check the container status in Docker or Portainer.
5. nextcloud-aio-fulltextsearch remains unhealthy.

Recovery steps already tried

I already tried the following recovery steps:

1. Stopped the AIO containers.
2. Removed all containers except the mastercontainer.
3. Started AIO again and let it recreate all containers.
4. The Fulltext Search container was still unhealthy.
5. Repeated the same procedure again.
6. Additionally removed the Fulltext Search Docker image.
7. Let AIO pull/download the Fulltext Search image again.
8. Started AIO again and let it recreate the container.

The issue still persists.

Recreating the containers did not help.
Removing and re-downloading the Fulltext Search image did not help either.

Relevant log output

The Fulltext Search container log contains the following messages:

Created elasticsearch keystore in /usr/share/elasticsearch/config/elasticsearch.keystore

GPU based vector indexing is not supported on this platform; Cannot create JDKProvider: libcuvs_c.so: libcuvs_c.so: cannot open shared object file: No such file or directory

SLF4J(W): No SLF4J providers were found.

SLF4J(W): Defaulting to no-operation (NOP) logger implementation

Transport SSL must be enabled if security is enabled. Please set [xpack.security.transport.ssl.enabled] to [true] or disable security by setting [xpack.security.enabled] to [false].

failed to retrieve password hash for reserved user [elastic]

Notes

I am not sure whether the GPU vector indexing and SLF4J messages are harmless Elasticsearch warnings, or whether they are related to the healthcheck failure.

The messages that look most relevant to me are:

• Transport SSL must be enabled if security is enabled
• failed to retrieve password hash for reserved user [elastic]

Since the issue appeared directly after updating to AIO 13.2.0 Beta / Nextcloud 33.0.5, and since it persists even after recreating the container and re-downloading the image, this may be related to the current beta Fulltext Search image, Elasticsearch configuration, or the container healthcheck.

I can provide the full nextcloud-aio-fulltextsearch container log and the output of the Docker healthcheck inspection if needed.

Screenshot

Logs attached.

_nextcloud-aio-fulltextsearch_logs.zip