EctoJob just suddenly stop.

[blisscs]

Hi,

We are experiencing that Ecto Job just stops working.

Our setup ->

• We have no. of elixir nodes (usually 3) set up to connect to one PostgreSQL database.

We continue to experience that Ecto Job just stop working. And we have to restart our elixir node to make things start working again.

When Ecto Job stops working, we can observe that there are unprocessed records (around 15-30 records) in AVAILABLE states in the job_queue table.

Please recommend how to check this issue and to protect this from happening again.

[mbuhot]

@blisscs this is the first report I’ve had if ecto_job “just stop working”.

Do you have any log output?

[blisscs]

@mbuhot ,

Sorry for the delay, I have no log of the issue.

What I could observe is that I have many network heavy tasks in the queue. And these network calls are usually timed longer than ecto multi's timeout.

I think with so many tasks like this, it eventually cause some effects to stop.