How to fix job stay queued problem?

[adarsh]

@tungxp

07/24/2019 11:25:47;0080;Server@hpc;Req;?;Req Header bad, errno 104, dis error 7
07/24/2019 11:25:47;0080;Server@hpc;Req;req_reject;Reject reply code=15056, aux=0, type=0, from @hpc.private
07/24/2019 11:25:47;0080;Server@hpc;Req;?;Req Header bad, errno 104, dis error 7
07/24/2019 11:25:47;0080;Server@hpc;Req;req_reject;Reject reply code=15056, aux=0, type=0, from @hpc.private

• The value of errno translates to “Connection reset by peer”. Suspect are something strange happening in the network or hostname resolution of the server . Maybe check for errors or dropped packets?
• protocol mis-match if you are using different versions of PBS Pro within your complex

@tungxp

PBS_LEAF_NAME=192.168.2.10
PBS_SERVER=192.168.2.10

You do not have to use PBS_LEAF_NAME here, please remove it.
PBS_SERVER=<hostname that is reverse address resolvable and same across all /etc/hosts of headnode and compute node >

Hostname issue setting up PBS Professional on CentOS 7

The value of PBS_LEAF_NAME should be the hostname of the interface over which you want PBS Pro to communicate, not the domain name. Unless you have multiple NICs, you should not need to set this value at all. Please try removing PBS_LEAF_NAME from /etc/pbs.conf and try starting PBS Pro.

@tungxp

#PBS -l nodes=1:ppn=1

This is old syntax, it is supported, better to switch to new syntax
#PBS -l select=1:ncpus=1