stargate's python process exits suddenly with pthread_cond_wait: Resource busy

[metamatt]

I left Stargate running for a few days and came back to find it had died with:

2013-02-22 07:32:23,368 dsc_listener WARNING : gateways.powerseries.dsc_gateway: ignoring action for partition:1
2013-02-22 07:36:33,396 dsc_listener WARNING : gateways.powerseries.dsc_gateway: ignoring action for partition:1
2013-02-22 07:40:43,424 dsc_listener WARNING : gateways.powerseries.dsc_gateway: ignoring action for partition:1
pthread_cond_wait: Resource busy
./stargate: line 3: 89212 Abort trap: 6           python ~/src/metamatt/stargate/stargate.py --config ./config.yaml

as its final words.

There was nothing additional in the logfile and no reason to think those powerseries messages were related; that's just context for what it had been up to recently.

I have no idea which condvar it was waiting on, which thread was waiting, or exactly what "resource busy" means to pthread_cond_wait.

[metamatt]

It happened again:

2013-03-05 04:03:21,663 db_cp_timer  WARNING : persistence: database checkpoint requested
pthread_cond_wait: Resource busy
./stargate: line 3:  6690 Abort trap: 6           python ~/src/metamatt/stargate/stargate.py --config ./config.yaml

[metamatt]

The host where I'm running this had limit coredumpsize 0. I unlimited it, so maybe that will yield useful information next time it happens.