[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] Master worker node problem

On Mon, Feb 20, 2006 at 08:37:34PM +0530, Srinivas Malyala wrote:
> Firsly i exicuted condor_master on worker node then after 1 minute i
> exicuted it on MASTER machine
> when i exicuted condor_status on master machine it showing only worker node
> but not master pc it (where i exicuted that command) for some time (say 1
> minute) after that it showing both.......
> why it happening like
> in general expectatation master should come first and then slave here why it
> reversed ???

The condor_startd does some benchmarking to measure the speed of the machine
before sending an update. The condor_collector process might very well start
before the condor_startd process on a machine, recieve an update message
from another startd, and answer a condor_status query before the startd
on the collector machine has finished it's benchmark. 

Bottom line: It's no big deal, don't worry about it :)