Jobs not being executed anymore??

[emas80]

Hi, upgrading the gem from 3.0.2 to 3.0.9 seems creating issue while executing some kinds of jobs.

I don't know how to build a simple test case to replicate it. It happens every time we are executing a particular job, but this job is not different from the others, and it doesn't even have the unique attribute so it should be totally ignored.
I mean, if the issue is present with this job, than it should be present with all the jobs...

The job contains an options object with { data => { match_id => "XXXXYYYYY"} }, I think something about the hashing didn't work properly because if one of them was enqueued when another one was already waiting to be executed (both using the "perform_in"), then the second was not executed, it was totally forgotten, like if it was never been added to the queue.

I downgraded to 3.0.2 and the issue is solved.

[mhenrixon]

Hmm for us it is the opposite. Before 3.0.10 job uniqueness didn't work. Is it possible that your redis hasn't been cleared up on a bunch of lock keys?

[RobinDaugherty]

Having the same issue. With a completely empty Redis instance, even though we're using mock_redis. Job arguments are a record ID that can't be a duplicate, because it's a record primary key. Using inline testing, the worker that has unique enabled never runs. I had to downgrade to 3.0.2, because any newer version fails consistently.

Also, since we're using mock_redis, the contents of any Redis instance should not matter, should they?

Please don't think that this issue is resolved; it seems that those like @emas80 have just locked their version to 3.0.2 and left it, like I'm about to do. We'll try to write up a test case.

[mhenrixon]

I'm not sure this problem will ever truly be solved. In our situation I am actually using the database to force uniqueness but in my case I need to never ever be able to re-run certain jobs.

Maybe mock redis was a bad idea... Anyway, give me a test case and lets fix it once and for all.

[mhenrixon]

Open a new issue with a failing test if this is still a problem.