doxout/recluster

Old workers still don't die!

codeaholics opened this issue · 3 comments

As with issue #1, I'm having difficulties getting old workers to die. On my dev machine, everything works fine. But in production when I send a SIGUSR2, new workers are started but the old ones remain. The old ones appear to continue doing work too.

What information/debugging can I provide to help solve this?

The default timeout for old workers is 1 hour. This is set so there is plenty of time for long file uploads to finish. You can make this shorter if you wish by seting opt.timeout to something else, e.g. 600 (10 minutes)

When NODE_ENV != production, the timeout is 1 second

edit: oh, apparently they're still doing work - are they still taking new requests too?

I added a test and new requests are only sent to new workers. Either that, or by some chance all of the 10 requests I send to the server arrive to the 2 new workers and none at the old.

Maybe your old workers are still working on old requests?

Fixed in v0.3.0