scheduler worker process looks frozen, stuck in state 'PICK'

[Brendan Fox]

So I've noticed that every few days my scheduler task appears to be frozen.  New tasks are put into the queue, but are not being processed.  When I look at the status of the single worker process in the system the db.scheduler_worker.status = 'PICK'.

I don't see this in the web2py docs anywhere.  What does this state mean and how could a scheduler process get stuck in this state?  The only way I've recovered is to restart the framework.

Thanks

[Dave S]

Roughly speaking, this indicates that you're in the middle of assigning a worker to a task (for a detailed answer, look in gluon/scheduler.py).

I have never seen my scheduler get stuck in that state, so I can't tell you much about it.  (The scheduler works very well for me, but I have simple needs.)


I think we need more information:

What OS is the server running on?
What database engine are you using?
What is the overall load on your server?
Are there error messages in the scheduler_run table?
How many queued tasks are shown in the scheduler_task table?
If you're not using the defaults, how many workers and how many groups?

/dps