Pydio Cells healthcheck endpoint does not appear to work

[jpts]

Describe your issue in detail

Got a new install of Pydio Cells v4 up and running, and trying to enable the healthcheck endpoint for monitoring.

Following this documentation, it seems to just create the env variable and it should enable the healthcheck endpoint; however, it doesn’t appear to be working.

Monitoring Tools

navigating my browser to http(s)://<ip-of-server>:6565/healthcheck returns nothing, as if there is no server running on port 6565. Reading the documentation, i expected a JSON response.

skimming the repo on Github, it seems the healthcheck flag has been commented out (?) - maybe i’m reading it wrong

Comment out unused flag for now

Here is a snippet of my docker-compse.yml

  cells:
    image: pydio/cells:latest
    restart: always
    ports: 
      - 8082:8080
      - 33060:33060
      - 6565:6565
    environment:
      - CELLS_BIND=${FQDN}:443
      - CELLS_SITE_BIND=${CELLS_SITE_BIND}
      - CELLS_LOG_LEVEL=${LOG_LEVEL}
      - CELLS_NO_TLS=${CELLS_NO_TLS}
      - CELLS_SITE_EXTERNAL=https://${FQDN}
      - CELLS_EXTERNAL=https://${FQDN}
      - CELLS_GRPC_EXTERNAL=33060
      - CELLS_HEALTHCHECK=6565
    volumes:
      - data:/var/cells/data
      - cellsdir:/var/cells

What version of Cells are you using?

Cells v4 Home Edition

What is the server OS? Database name/version? Browser name or mobile device description (if issue appears client-side)?

running in Docker on top of Unraid.

[bsinou]

Hi and welcome in our community!

So, indeed the dedicated health check entry point (as described in the doc you are pointing to) has been broken for a while now.

On our side, typically in docker setups, we rather rely on a GET call on the login page:
<your FQDN>/login
It returns an HTTP status 200 when the server is up and running.

We removed the flag in latest release but forgot to update the doc, we’ll do it soon. Thanks for the pointer!

[jpts]

thanks @bsinou for the confirmation. That was my guess after reading the repo.

checking the login page is what i’ve currently doing, and what i’ll continue to use for a healthcheck