[slurm-users] Slurm 19.05: can not submit job

[Ole Holm Nielsen]

On 11/28/19 11:47 AM, Nguyen Dai Quy wrote:
> 
> On Thu, Nov 28, 2019 at 11:20 AM Ole Holm Nielsen 
> <Ole.H.Nielsen at fysik.dtu.dk <mailto:Ole.H.Nielsen at fysik.dtu.dk>> wrote:
> 
>     On 11/28/19 10:35 AM, Nguyen Dai Quy wrote:
>      > Hi list,
>      > I can not submit my job:
>      >  > sbatch submit.sh
>      > sbatch: error: Batch job submission failed: Invalid account or
>      > account/partition combination specified
>      >
>      > After checking slurmdbd.log, I see:
>      >
>      > [2019-11-28T10:21:07.578] Accounting storage MYSQL plugin loaded
>      > [2019-11-28T10:21:07.586] slurmdbd version 19.05.4 started
>      > [2019-11-28T10:26:07.778] error: _add_registered_cluster: trying to
>      > register a cluster (cluster3) with no remote port
>      > [2019-11-28T10:30:14.936] Terminate signal (SIGINT or SIGTERM) received
>      > [2019-11-28T10:30:14.951] Unable to remove pidfile
>      > '/var/run/slurmdbd.pid': Permission denied
>      > [2019-11-28T10:30:15.038] Accounting storage MYSQL plugin loaded
>      > [2019-11-28T10:30:15.047] slurmdbd version 19.05.4 started
>      > [2019-11-28T10:31:07.997] error: _add_registered_cluster: trying to
>      > register a cluster (cluster3) with no remote port
>      >
>      > I used slurm 19.05 on CentOS 7.
>      > Any suggestions?
> 
>     This could perhaps be a firewall problem.  I suggest that you try to
>     validate your configuration by following my Slurm Wiki:
>     https://wiki.fysik.dtu.dk/niflheim/SLURM
> 
>     In particular, you have to configure the CentOS 7 firewall correctly:
>     https://wiki.fysik.dtu.dk/niflheim/Slurm_configuration#configure-firewall-for-slurm-daemons
>     https://wiki.fysik.dtu.dk/niflheim/Slurm_configuration#firewall-between-slurmctld-and-slurmdbd
> 
>     /Ole
> 
> 
> Thank you for you reply.
> We do not have a firewall. My problem has been corrected.
> The cause is bad directory permission of SlurmdSpoolDir=/var/spool/slurmd
> After correction, it works well.

Nice.  However, CentOS 7 is running the firewalld by default, and Slurm 
works well with the Linux firewall, but you have to configure it correctly 
as shown in the above Wiki pages.

/Ole