NIC driver or what?
-
I'm on latest on HEAD: 2.1-BETA1 (i386) built on Sun May 5 07:29:51 EDT 2013
This is my setup:
–------------Internet---------------
^
|
------------------------
| |
ADSL modem 3g Modem
| |
| |
NIC 1(fxp2) WAN NIC 2(fxp3) OPT1
| |
----------pfsense--------
|
NIC 3(bge0) LANI only enable fxp3 when ADSL goes down (3g is a time based contract, so cannot do failover).
The problem:
Every morning I find NIC 1 down, with interface status bouncing between "up" and "no carrier".
Log entries:May 6 11:56:37 check_reload_status: Linkup starting fxp2 May 6 11:56:35 kernel: fxp2: link state changed to DOWN May 6 11:56:35 check_reload_status: Linkup starting fxp2 May 6 11:56:35 php: : HOTPLUG: Configuring interface wan May 6 11:56:35 php: : DEVD Ethernet attached event for wan May 6 11:56:33 php: : DEVD Ethernet detached event for wan May 6 11:56:32 kernel: fxp2: link state changed to UP
This loops continuously, with CPU on 100% with check_reload_status, until eventually internet access goes down completely, with no particular log entry.
I started having this problem with the snapshots from april…
I tried everything: changing cable, changing NICs with brand new ones, even changed hard drive as SMART status was not looking too good.Today ADSL went down, so I enabled OPT1 (fxp3). Started having the same identical problem with that interface:
May 6 11:56:37 check_reload_status: Linkup starting fxp3 May 6 11:56:35 kernel: fxp3: link state changed to DOWN May 6 11:56:35 check_reload_status: Linkup starting fxp3 May 6 11:56:35 php: : HOTPLUG: Configuring interface opt2 May 6 11:56:35 php: : DEVD Ethernet attached event for opt2 May 6 11:56:33 php: : DEVD Ethernet detached event for opt2 May 6 11:56:32 kernel: fxp3: link state changed to UP
Everything functions ok after a restart, so I setup a cron job to "shutdown -r now" every early morning, but this is clearly a hack, and restarting during work hours is a no-no.
Have the drivers been recently modified? What am I missing?
I'm really desperate, please give me hand on finding this bugger…
-
-
Yeah, I think we should create dedicated thread and test things out…
I˙m willing to cooperate fully to resolve this issue or bug if it trully exists...