Another stack trace

[Dave Abrahams]

on Tue Sep 27 2011, Sebastian Spaeth <Sebastian-AT-SSpaeth.de> wrote:

> On Mon, 26 Sep 2011 15:17:23 -0400, Dave Abrahams <dave at boostpro.com> wrote:
>> FWIW, I think what we were seeing was Python's response to the process
>> being SIGKILLed, which I do periodically to get it to start a fresh
>> sync.  It'd be nice if the process would respond to SIGHUP, though :-)
>
> I think what you want it to use the SIGUSR1 and SIGUSR2 signals to start
> a new round of syncs?!

Do you mean that those signals will start a new round of syncs today, or
that you think I should be asking for something other than SIGHUP,
or...?

> And no, I am not too fond of the choice of signals. SIGHUP and/or
> SIGCONT would make a lot more sense to start a new round of refresh
> runs, IMHO.

What's the difference between a refresh run and a sync?

-- 
Dave Abrahams
BoostPro Computing
http://www.boostpro.com