Hello again,
Any help in this topic would be appreciated. Thank you.
El 19/05/15 a las 17:56, Elena Núñez González escribió:
Hello,
I'm working with the SLURM simulator in a Centos7 machine, so I'm
working in the branch simulator from the GitHub repository. I've
compiled and installed the simulator.
When I execute
| # su - slurm
$ /home/SLURM_SIM/sbin/slurmctld -Dcv
$ /home/SLURM_SIM/sbin/slurmd -Dcv
$ /home/SLURM_SIM/bin/sinfo|
It works. Then, following the instructions, I executed
|$ cd /home/SLURM_SIM/sbin
$ ./exec_sim.pl SIM_DIR 100|
Everytime, after a while I get this error by the command line:
slurm_load_jobs error: Unable to contact slurm controller (connect
failure)
But, anyway it looks like it is working, because the log exec_sim.log
tell, at the end:
Getting maps for process ID 31672
Waiting...
Ok. We have 1000 completed jobs
Killing simulation processes...
In the documentation they talk about the log 'tmp/job_comp.log', and I
don't have it.
But, when I try to execute
$ sbin/list_trace
I get this error, and nothing happens:
Error opening test.trace
I've been looking for the file, and it doesn't exist.
I don't know if there is some mistake, and how can I work with the
simulator. Any help?
I'll give you some more information, maybe it helps.
The first time I run the command
sbin/slurmdbd -Dvv
Everything looks fine, but the next times, it gives the following
messages for every user in the Database:
slurmdbd: debug: post user: couldn't get a uid for user par11040
In the log slurmd.log I get the same kind of messages:
[2015-05-19T11:27:22.194] debug: post user: couldn't get a uid for
user tra10015
[2015-05-19T11:27:22.194] slurmdbd version 15.08.0-0pre3 started
But in every case, the process says it is started.
If anybody has any advice, i would really appreciate it.
Thank you
