Hi
I am seeing a lot of errors like that in log file at our torque server
pbs_server: LOG_ERROR::Cannot assign requested address (99) in send_job, send_job failed to a301058e port 15002
where 130105* are IP addresses in hexadecimal format. It started after we updated to torque-server-2.5.7-7. It is random and intermittent. Probably because of this some time jobs goes into W state or Q state with a node assign to it. Is some one else also seeing it ?
Thanks
Kashif
|