[bug]: context_chat_backend never finishes heartbeat check (CPU/AVX2)

[kds69]

context_chat_backend never finishes heartbeat check, but app seems to have started up.

Steps to reproduce the behavior:
One Click deployment against default docker install

Context Chat Backend logs (if applicable, from the docker container)

Starting PostgreSQL...
sudo: unable to resolve host context_chat_backend: Name or service not known
pg_ctl: another server might be running; trying to start server anyway
waiting for server to start.... done
server started
Waiting for PostgreSQL to start...
PostgreSQL is up and running.
sudo: unable to resolve host context_chat_backend: Name or service not known
sudo: unable to resolve host context_chat_backend: Name or service not known
ALTER ROLE
sudo: unable to resolve host context_chat_backend: Name or service not known
sudo: unable to resolve host context_chat_backend: Name or service not known
NOTICE:  extension "vector" already exists, skipping
CREATE EXTENSION
Running repairs...
No repairs are required.
Config file already exists in the persistent storage ("/nc_app_context_chat_backend_data/config.yaml").
App config:
{
  "debug": true,
  "uvicorn_log_level": "info",
  "disable_aaa": false,
  "httpx_verify_ssl": true,
  "use_colors": true,
  "uvicorn_workers": 1,
  "embedding_chunk_size": 2000,
  "doc_parser_worker_limit": 10,
  "vectordb": [
    "pgvector",
    {}
  ],
  "embedding": {
    "protocol": "http",
    "host": "localhost",
    "port": 5000,
    "workers": 1,
    "offload_after_mins": 15,
    "request_timeout": 1800,
    "llama": {
      "model": "multilingual-e5-large-instruct-q6_k.gguf",
      "n_batch": 16,
      "n_ctx": 8192
    }
  },
  "llm": [
    "nc_texttotext",
    {}
  ]
}
INFO:     Started server process [45]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:23000 (Press CTRL+C to quit)

Screenshots

Setup Details (please complete the following information):

• Nextcloud Version: 30.0.5
• AppAPI Version: 4.0.5
• Context Chat PHP Version N/A (not installed yet)
• Context Chat Backend Version 4.0.7
• Nextcloud deployment method: Apache VM (proxmox)
• Context Chat Backend deployment method: one-click

[kyteinsky]

hello, this is an issue with app_api setup. Transferring it to app_api's repo for its resolution there.

would you mind describing your setup a little? The daemon configuration, if it's a remote setup (different server for ex-apps) and most importantly, the network config.

[uncaney]

any update on that issue ? it also appear that it appear for a few seconds in docker ps then stop and disapear.

[uncaney]

i get that kind of error (either with context chat backend or the agent)

Apr 18 20:22:48 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:22:48.034719493Z" level=info msg="Container failed to exit within 10s of signal 15 - using the force" container=73c3ecc98bb78b4efb197ad7dcc1278a4d524782601f5160ca835aafc95d6afe
Apr 18 20:22:48 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:22:48.119692481Z" level=info msg="ignoring event" container=73c3ecc98bb78b4efb197ad7dcc1278a4d524782601f5160ca835aafc95d6afe module=libcontainerd namespace=moby topic=/tasks/delete type="*events.TaskDelete"
Apr 18 20:22:48 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:22:48.140106668Z" level=warning msg="ShouldRestart failed, container will not be restarted" container=73c3ecc98bb78b4efb197ad7dcc1278a4d524782601f5160ca835aafc95d6afe daemonShuttingDown=false error="restart canceled" execDuration=13.376671022s exitStatus="{137 2025-04-18 20:22:48.100508385 +>
Apr 18 20:28:33 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:28:33.049772324Z" level=info msg="Attempting next endpoint for pull after error: manifest unknown"
Apr 18 20:28:33 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:28:33.052931498Z" level=error msg="Handler for POST /v1.41/images/create returned error: manifest unknown"
Apr 18 20:28:33 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:28:33.334896658Z" level=info msg="Attempting next endpoint for pull after error: Head "https://ghcr.io/v2/nextcloud/context_agent-cpu/manifests/1.1.0\": denied"
Apr 18 20:28:33 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:28:33.337700067Z" level=error msg="Handler for POST /v1.41/images/create returned error: Head "https://ghcr.io/v2/nextcloud/context_agent-cpu/manifests/1.1.0\": denied"
Apr 18 20:29:01 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:29:01.608992280Z" level=info msg="Container failed to exit within 10s of signal 15 - using the force" container=80d2037f45c89e0a6e82a2e8f4f92153c1c9fcc2bb0f30d69eafe8296c6037e8
Apr 18 20:29:01 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:29:01.653728843Z" level=info msg="ignoring event" container=80d2037f45c89e0a6e82a2e8f4f92153c1c9fcc2bb0f30d69eafe8296c6037e8 module=libcontainerd namespace=moby topic=/tasks/delete type="*events.TaskDelete"
Apr 18 20:29:01 ubuntu-nextcloud-nbg1-1 dockerd[79099]: time="2025-04-18T20:29:01.672332036Z" level=warning msg="ShouldRestart failed, container will not be restarted" container=80d2037f45c89e0a6e82a2e8f4f92153c1c9fcc2bb0f30d69eafe8296c6037e8 daemonShuttingDown=false error="restart canceled" execDuration=13.812793673s exitStatus="{137 2025-04-18 20:29:01.63747477 +0>

i have to pull it myself, restart docker then enable it manually

[kyteinsky]

hello,
can you let us know if "test deploy" works for you? It should be in the 3 dots menu in the daemon admin settings.

if yes, can you provide the docker container's logs docker logs nc_app_context_chat_backend and the server logs for the time the installation was going down?

PS. the image pull errors containing "cpu" is just app_api checking if there is a specialised cpu image available, it's normal.

[oleksandr-nc]

We are closing this issue as obsolete. If you still care about this, we can recommend upgrading to Nextcloud 31 and creating a new issue for the test-deploy action where we can try to triage problem.
If you can wait, in Nextcloud 32 we will provide an alternative to DockerSocketProxy - which is easier to configure.