MT urgent problem – I have upgraded many times and isolated the problem not to be with my interface or switches or anything else but seems like MT. The intel nic we use since long time is 82546GB (rev 03) Please see the 2.9.25 rif and earlier rifs i have sent. The interface is dual port nic on pci-x. Both the ports stop forwaarding traffic suddenly just nothing. My switches show the interface as up and running . THERE IS NO TRAFFIC ON THE INTERFACES.
At this time I have switched machines, changed nics with 3 others and tried everythign but this is random. The nics if i enable disable them they come back up but after some time just stop forwarding traffic. I even changed the ports from gigE to 100mbps to switch changes. NO USE
Can you please go through supout urgently and let me know whether you see anything . I have upgraded to 2.9.26 anyway to see if tht helps (though after going through 2.9.19, 2.9.24 and 2.9.25 over the last 3 weeks I am pretty much sure there is something).
Have reached a point where few more downtimes and will have to junk MT for something else.
Are these 2 nics on the same physical segment? I wonder if this is part of the problem. Also, if you just use 1 nic does it work well? From the readme on intels site about nics on the same segment:
Multiple Interfaces on Same Ethernet Broadcast Network
Due to the default ARP behavior on Linux, it is not possible to have
one system on two IP networks in the same Ethernet broadcast domain
(non-partitioned switch) behave as expected. All Ethernet interfaces
will respond to IP traffic for any IP address assigned to the system.
This results in unbalanced receive traffic.
If you have multiple interfaces in a server, either turn on ARP
filtering by entering:
echo 1 > /proc/sys/net/ipv4/conf/all/arp_filter
(this only works if your kernel’s version is higher than 2.4.5),
NOTE: This setting is not saved across reboots. The configuration
change can be made permanent by adding the line:
net.ipv4.conf.all.arp_filter = 1
to the file /etc/sysctl.conf
or,
install the interfaces in separate broadcast domains (either in
different switches or in a switch partitioned to VLANs).
The nic is a single dual port pro 1000mt server adapter on pci-x
one port talks to one provider while the other talks to another provider. All this over separate vlans on a cisco 2970 and now a 2950(to make sure it wasnt the switch) . I have this running for over 3 years now (mt 2.8.x onward), this problem has recently started 3 weeks ago and im still looking for answers.