[AG-TECH] VenueServer lockup

[Kaizaad Bilimorya]

VenueServer running AccessGrid: 3.0.2
Fedora Core 6 with Kernel: Linux agvs.sharcnet.ca 2.6.22.9-61.fc6 #1 SMP Thu Sep 27 18:48:03 EDT 2007 i686 athlon i386 GNU/Linux

Seems that our VenueServer locks up occasionally:

[root at agvs Logs]# strace -p 25297
Process 25297 attached - interrupt to quit
futex(0xa1ef168, FUTEX_WAIT, 0, NULL
...hangs here

The last entries in the Log don't seem to have anything obvious:

10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:239 DEBUG Client Timeout Check: 1193322246 vs 1193322136
10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:685 DEBUG CleanupClients: now=1193322136
10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:685 DEBUG CleanupClients: now=1193322136
10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:685 DEBUG CleanupClients: now=1193322136
10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:685 DEBUG CleanupClients: now=1193322136
10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:685 DEBUG CleanupClients: now=1193322136
10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:685 DEBUG CleanupClients: now=1193322136
10/25/07 10:22:16 -1241420912 VenueServer     Venue.py:685 DEBUG CleanupClients: now=1193322136


This seems to happen more when we have people connected on our bridge, 
though not always and not reproducible.

What is the VenuServer trying to do here? Contact each client and make 
sure they are alive?

Any debugging hints would be greatly appreciated.

thanks
-k