[torqueusers] mount.nfs starts failing after pbs_server gets "warmed up", starts using up more than 700 sockets (Sabuj Pattanayek)Re:

Mark Henshall Mark.Henshall at cancer.org.uk
Mon Apr 22 13:51:56 MDT 2013

That worked a treat - I was slightly worried about the 
net.ipv4.tcp_tw_recycle kernel change, but the doc I found that in a 
couple of weeks ago was talking about http servers.

The --disable-privports is the answer. A user has kicked in recently 
using mpi - on Friday, nfs was a big problem - now, no problems at all.

Thanks everyone.

Quoting Sabuj Pattanayek <sabujp at gmail.com>:

>> but I reckon I'm going to have to do a recompile - will the reconfig
>> mean I have to push out the pbs_mom binary to the nodes (I can see that
>> being an issue), or is it just the pbs_server binary that'll need
>> replacing and restarting? If it's the latter, then great.
> Just the latter since pbs_server is making a connection per job and
> would use up all ports < 1024 which mount.nfs also tries to use
> without that configure flag. I'm seeing that pbs_mom is already using
> ports > 1024 for the most part:
> tcp        0      0     *
>     LISTEN      2492/pbs_mom
> tcp        0      0     *
>     LISTEN      2492/pbs_mom
> udp        0      0      *
>                 2492/pbs_mom
> udp        0      0     *
>                 2492/pbs_mom
> which is fine and won't cause any issues with nfs mounting on the nodes.
> HTH,
> Sabuj
> _______________________________________________
> torqueusers mailing list
> torqueusers at supercluster.org
> http://www.supercluster.org/mailman/listinfo/torqueusers

Mark Henshall
Unix Administrator
Cancer Research UK London Research Institute
Lincoln's Inn Fields Laboratories
44 Lincoln's Inn Fields
London WC2A 3LY
Registered charity number 1089464

t: 0207 269 3602
m: 07890664727
f: 0207 061 8011
e: Mark.Henshall at cancer.org.uk

This e-mail (including any attachments) is intended for the above-named person(s). If you are not the intended recipient, notify the sender immediately, delete this email from your system and do not disclose or use for any purpose. 

We may monitor all incoming and outgoing emails in line with current legislation. We have taken steps to ensure that this email and attachments are free from any virus, but it remains your responsibility to ensure that viruses do not adversely affect you. 
Cancer Research UK
Registered charity in England and Wales (1089464), Scotland (SC041666) and the Isle of Man (1103)
A company limited by guarantee.  Registered company in England and Wales (4325234) and the Isle of Man (5713F).
Registered Office Address: Angel Building, 407 St John Street, London EC1V 4AD.

More information about the torqueusers mailing list