Re: [HTCondor-users] [PATCH] Speeding up condor_dagman submission

["R. Kent Wenger" <wenger@xxxxxxxxxxx>]

On Mon, 3 Aug 2015, Brian Candler wrote:

> Hmm, deathly silence. Would anyone in the Condor core team care to take a 
> look at this, and/or explain the logic behind the sleeps in the algorithm for 
> assigning process ids?
>
> References:
> https: //www-auth.cs.wisc.edu/lists/htcondor-users/2015-July/msg00021.shtml
> https: //www-auth.cs.wisc.edu/lists/htcondor-users/2014-December/msg00053.shtml
>
> It's amazing how much pain a 12-second delay can cause when it's repeated 
> over and over again :-(

Sorry for not replying to this earlier...

Anyhow, we're investigating whether we can reduce (or entirely remove) the 
12-second sleep without running into problems.  I think the conversions 
between seconds and clock ticks *are* consistent between the various code 
paths; but the real question is whether we need the 12-second sleep.

It's certainly a bug, though, that the value of DEFAULT_PRECISION_RANGE 
will get changed if the code gets called twice.  (That shouldn't actually 
be causing a problem right now, because we only call the relevant code on 
startup, but it still should be fixed.)

(This code was written by someone who's no longer with the CHTC Team, and 
unfortunately it's not real well understood by any of the current team 
members.)

Kent Wenger
CHTC Team