On Sat, Jun 8, 2013 at 8:36 PM, Jason Pyeron <[email protected]> wrote:

> **
> The only time I have observed that type of problem was when the power
> supply was browning out, ensure that your power supply is of a good quality
> and sufficient amperage.
>


Thanks for the input. My power supply is actually a DC setup with redundant
switching power supplies and battery backup. I have 8 modems, 2 PoE radios
and a Core i3 server with a DC PSU all powered from the same setup, and all
running apparently fine when pfsense locks up, so I'm reasonably confident
that the power is ok in this case.

I took the time to filter and arrange my 'netstat -m' logs into a Google
spreadsheet. It's a limited amount of data, but I think it at least
suggests a correlation between exhausted 9k jumbo clusters and the
lock-ups. Whether this correlation is significant, I don't know. I'm hoping
somebody on this list does.

https://docs.google.com/spreadsheet/ccc?key=0AvXA5Dt3xhI-dG9LblgtOTh2ZmQ5VTV6d1hDaVNFZGc&usp=sharing

Meanwhile, I had a 9k mtu set on the LAN interface. I have set that to the
default 1500 like all the others in hope that this will stop the problem
while I research the significance of the 9k clusters that I'm seeing
inflate.

db
_______________________________________________
List mailing list
[email protected]
http://lists.pfsense.org/mailman/listinfo/list

Reply via email to