Re: [Condor-users] Problems with jobs

["Chris Miles" <chrismiles@xxxxxxxxxxxxxxxx>]

Basically I have a pool with a shared file system and 25 machines. These are 
all very powerfull.
I think the weakest link in my chain is my submitting machine which is just 
a lone server with its
own configuration. Its a 1ghz 512mb Mini ITX box. Not the fastest in the 
world and has a few
other applications running (required).

Is this the machine I should set JOB_START_COUNT on? or should it be set on 
the machines that
actually run the jobs?

On my submitting machine thats the one I see the condor_shadow daemons 
firing up.


chris    18253  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.9 
<146.191.100.202:46251> -
chris    18325  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.12 
<146.191.100.202:46251> -
chris    18362  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.13 
<146.191.100.202:46251> -
chris    18396  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.14 
<146.191.100.202:46251> -
chris    18454  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.15 
<146.191.100.202:46251> -
chris    18464  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.16 
<146.191.100.202:46251> -
chris    18499  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.17 
<146.191.100.202:46251> -
chris    18533  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.18 
<146.191.100.202:46251> -
chris    18570  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.19 
<146.191.100.202:46251> -
chris    18579  5552  0 14:01 ?        00:00:00 condor_shadow -f 80.20 
<146.191.100.202:46251> -

How many of these is normal? I am submitting a 1000 job cluster to the pool 
with 25 machines (50 vms).

Looks like I may be running low on memory on my submitting machine as well.

top - 14:01:52 up 20 days, 21:24,  3 users,  load average: 0.79, 1.21, 0.92
Tasks:  71 total,   1 running,  70 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.3% us,  1.7% sy,  0.7% ni, 97.0% id,  0.0% wa,  0.3% hi,  0.0% si
Mem:    484284k total,   476868k used,     7416k free,    14216k buffers
Swap:   999928k total,        0k used,   999928k free,   249768k cached

Im still unsure to some of this.... where exactly is the problem lying,, the 
submitter or the executers?

thanks again

Chris

----- Original Message ----- 
From: "Matt Hope" <matthew.hope@xxxxxxxxx>
To: "Condor-Users Mail List" <condor-users@xxxxxxxxxxx>
Cc: "Ian Chesal" <ICHESAL@xxxxxxxxxx>
Sent: Thursday, December 08, 2005 8:21 AM
Subject: Re: [Condor-users] Problems with jobs


> On 12/7/05, Chris Miles <chrismiles@xxxxxxxxxxxxxxxx> wrote:
> > I have managed to get that number up as high as 20 and even 50 with only
> > little difference. I am seeing
> > more running jobs, but not much more. Only 7vms max so far
>
> How many (non held) clusters and jobs* are in your queue and how often
> do you negotiate?
>
> Since the schedd can only do one of the two tasks (starting shadows
> and serving queue info requests) it can fail to keep up
>
> A similar situation can occur if something/someone is running condor_q
> against your schedd repeatedly.
>
> * if NEGOTIATE_ALL_JOBS_IN_CLUSTER is true then jobs matter, if not
> then clusters matter.
>
> Matt
>
> _______________________________________________
> Condor-users mailing list
> Condor-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/condor-users