Folks;
I give credit where credit is due, and I thank Brian K. West
What For:
This was found to be a compounded problem. (Cisco was part of it...
But the real problem was the linux kernel...)
Suffice it to say, without the kernel bug, the cisco bug wouldn't have
been easily found.
What Kernel Bug:
It's a kernel bug that corrupted the sqlite database.
This caused Freeswitch to refuse the phones registration request.
This in turn caused the phones to re-register.
Problem was, with 10 phones, 6 lines each, perpetually registering on
a 100Mbps LAN, well, you can imagine the overhead.
This created severe latency with Freeswitch, and manifested as dropped
calls, one way audio, and the more phones you had, the worse the
problem was.
Workaround for problem was to use a ramdisk (tmpfs) for the database
- (Big Thanks to bkw!)
So far, it's been 24 hours, and all systems are nice and stable. (no
negative reports yet (fingers crossed)).
Brian (and Folks);
If this is stable through Friday (and there's no reason to think it
won't be), I will take the time to document the problem, basic
configuration, and the workaround for the problem on the Wiki this
weekend.
Best Regards,
Karl J. Vesterling
k...@ken-ton.com
202-461-3231 x0
On Sep 14, 2009, at 9:22 AM, Brian West wrote:
HAHA I couldn't have said this better!
/b
On Sep 14, 2009, at 8:17 AM, Anthony Minessale wrote:
The first hint was when the firmware rev began with the letters POS
_______________________________________________
FreeSWITCH-users mailing list
FreeSWITCH-users@lists.freeswitch.org
http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-
users
http://www.freeswitch.org
_______________________________________________
FreeSWITCH-users mailing list
FreeSWITCH-users@lists.freeswitch.org
http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
http://www.freeswitch.org
