How to resolve ‘state = maintenance’ node

Users/Site Administrators
1

Hi evrybody,

I have a strange problem

After an overloading on network I use qsig -s suspend jobid .
The job was job_array ( array size=3279 )

Pbs set my concerned nodes “maintenance” ( no problem )

after that our network was always overloaded then we kill jobs

with qdel command

but since the qdel we have node flagged “maintenance” that won’t return to free state

I tried pbsnodes -r node , and
qmgr -c “set node node state = free”

but nodes remain in “maintenance” mode

Have some Ideas ?

pbs_version = 20.0.0
debian 11

Best regards

2

Hi everybody

I found a manner

del node nodexx

cre node nodexx
then set all old config

and node are ok

He someone has better ??

Best Regard

3

Please share the output of the below commands when the node is in maintenance mode

pbsnodes  nodename-of-the-maintaince-node
qstat -Q
pbs_rstat
4

Hi,

Thank you for your answer … but now my prob is resolved …

The “deleting recreating method” works perfectly

Best Regards

1 Like