how to test crashsms about superlance

[mnaberez]

crashsms only sends an alert when it receives the PROCESS_STATE_EXITED event, meaning the process has entered the EXITED state. When you stop the process with supervisorctl stop, it puts the process into the STOPPED state. Try causing the process to exit another way, e.g. by kill -9 <pid> or since Supervisor 3.3.0, supervisorctl signal 9 <processname>. To see detailed messages about what events are being sent to event listeners like crashsms, set loglevel=blat in the [supervisord] section of the config file.

[morfin]

Good, I killed it and this is the log (Still not getting any email):

worker-1000:worker-1000 exited unexpectedly 2016-10-18 16:29:00,313

2016-10-18 16:29:00,326 DEBG event 5 sent to listener crashsms
2016-10-18 16:29:00,326 BLAT read event caused by <PEventListenerDispatcher at 140586332972960 for <Subprocess at 140586333223392 with name crashsms in state RUNNING> (stdout)>
2016-10-18 16:29:00,326 DEBG 'crashsms' stdout output:
RESULT 2
OKREADY

2016-10-18 16:29:00,327 DEBG crashsms: event was processed
2016-10-18 16:29:00,327 DEBG crashsms: BUSY -> ACKNOWLEDGED
2016-10-18 16:29:00,327 DEBG crashsms: ACKNOWLEDGED -> READY
2016-10-18 16:29:01,329 INFO success: worker-1000 entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)

Any other idea? Is there any log where I can see how crashsms sends the message?

By the way, loglevel=blat didn't work. loglevel=blather worked.

[mnaberez]

crashsms doesn't have a log file. It might print some messages to its stderr, which you can see with supervisorctl tail <processname> stderr or in the main supervisord log if the log level is low enough (like blat).

We use this issue tracker mainly for bugs. The supervisor-users mailing list is a good place to ask questions installing or using Supervisor and Superlance.