[Condor-users] Submit test wastes a long long time

[Fabiano Portella <fabiano_portella@xxxxxxxxxxxx>]

Hi Condor folks!

 

I've setup Condor in 6 machines of a lab. I've done that through SSH (I'm working from home). I can see status of just 5 of them, but this is not the case (maybe the 6th one was turned off).

Well, after this step, I started to test this pool by submitting 2 jobs from different nodes.

 

One test is the env.C code from "examples" folder

The other is from http://www.cs.wisc.edu/condor/tutorials/fermi-2005/submit_first.html page (simple.c code).

 

Although both are simple tests, they took a long time to finish (something about 30 minutes in first test and more than an hour in second test).

Following is the env.log file:

 

======================================================

 

000 (001.000.000) 04/06 15:09:59 Job submitted from host: <IP:PORT>
...
001 (001.000.000) 04/06 15:30:03 Job executing on host: <IP:PORT>
...
006 (001.000.000) 04/06 15:30:03 Image size of job updated: 2532
...
003 (001.000.000) 04/06 15:30:03 Job was checkpointed.
        Usr 0 00:00:00, Sys 0 00:00:00  -  Run Remote Usage
            Usr 0 00:00:00, Sys 0 00:00:00  -  Run Local Usage
...
003 (001.000.000) 04/06 15:30:03 Job was checkpointed.
        Usr 0 00:00:00, Sys 0 00:00:00  -  Run Remote Usage
            Usr 0 00:00:00, Sys 0 00:00:00  -  Run Local Usage
...
005 (001.000.000) 04/06 15:30:04 Job terminated.
        (1) Normal termination (return value 0)
            Usr 0 00:00:00, Sys 0 00:00:00  -  Run Remote Usage
            Usr 0 00:00:00, Sys 0 00:00:00  -  Run Local Usage
            Usr 0 00:00:00, Sys 0 00:00:00  -  Total Remote Usage
            Usr 0 00:00:00, Sys 0 00:00:00  -  Total Local Usage
        1463185  -  Run Bytes Sent By Job
        1864261  -  Run Bytes Received By Job
        1463185  -  Total Bytes Sent By Job
        1864261  -  Total Bytes Received By Job
...

======================================================

 

As you can see, the job started just 21 minutes after submitted.

My question is: is it possible that I need to change some configuration setting or this can only be caused by other concurrent process with greater priority in node?

Thanks in advance.

Regards,

Fabiano.

---

Abra sua conta no Yahoo! Mail - 1GB de espaço, alertas de e-mail no celular e anti-spam realmente eficaz.