<ping> can anybody comment on the internals?
i can dig through the code, but for anybody that's familiar with the internal state machine of supervisord, this might be a super easy question to answer. On Thu, May 17, 2012 at 1:50 PM, David Birdsong <[email protected]> wrote: > will startsecs block the entire supervisor loop while starting up a > series of programs? > > here's a scenario i'm trying to fix(psuedocode-ish). > > [program1] > priority=100 > command=mongos > > [program2] > priority=999 # default > command=java # running a jetty > > > the program2 is invoked after the program1 due to priorities, but it's > not long enough, the rapid succession of forking mongos then a jetty > means that the jetty fails to connect to a 'sill-initializing' mongos > and fails to boot up and generally mucks up our running status into a > bloody mess. > > the default startsecs == 1. > - does the succession from program1 to program2 block for that 1 second? > - if i raise it to say 4, will supervisor fork/exec program1, wait 4 > seconds check it's for an unexpected exit, put program1 into RUNNING > state and only then continue on to start program2? _______________________________________________ Supervisor-users mailing list [email protected] http://lists.supervisord.org/mailman/listinfo/supervisor-users
