10-08-2014 04:37 AM - edited 03-01-2019 11:52 AM
Hello,
I recently upgraded my Stand-Alone C22 M3 blade from version 1.5.5 to version 2.0(3d)
The VIC 1225 got firmware version 4.0(1b) (old version 2.2(1b))
After the upgrade using the Host Upgrade Utility I removed all old drivers and installed the newest versions.
so far so good, but almost immediately i noticed very poor basic nateworking.
We use this blade as a commvault backup server and both VIC 1225 Nics are in a LACP team (Status is Active)
on the host switch Nexus 5548UP there is a FC interface bind to the network interface so we have our backup libaries available on the host
Only if i keep pinging the servers i want to backup the network connection stays stable, otherwise backups will fail.
after downgrading the firmware of the vic back to 2.2(1b) everything went to normal again.
is there any explanation for this? known bug?
Thanks,
Ivo
Solved! Go to Solution.
10-08-2014 06:58 AM
You also have a etherchannel configured on your nexus 5K, isn't it right?
Take a look at the logging messages and try to find a MAC_FLAPPING message related to the servers MAC addresses.
Also check out this bug: https://tools.cisco.com/bugsearch/bug/CSCuf65032.
It was corrected on the version 2.22(1b) (your old one), but it may be back on the 4.0(1b).
Try to apply the workaround suggested and see if it works.
Cheers,
Kauy Gabriel.
10-08-2014 05:36 AM
Hi Ivo.
That loss of connectivity might be related with the OS.
Have you checked for updates on the OS' NIC drivers after upgrading?
If you also have a NIC teaming I would recommend you to recreate that too.
Cheers,
Kauy Gabriel.
10-08-2014 06:25 AM
Hi Kauy,
I tried the drivers windows update recommended me (very old (2013)), no improvement
I tried different versions from the cisco site, no improvement
Recreating the NIC Team with all possible configurations LACP, Static Switch and Independent Switch, no improvement.
the only thing that helped was downgrading the VIC firmware
Kind regards,
Ivo
10-08-2014 06:58 AM
You also have a etherchannel configured on your nexus 5K, isn't it right?
Take a look at the logging messages and try to find a MAC_FLAPPING message related to the servers MAC addresses.
Also check out this bug: https://tools.cisco.com/bugsearch/bug/CSCuf65032.
It was corrected on the version 2.22(1b) (your old one), but it may be back on the 4.0(1b).
Try to apply the workaround suggested and see if it works.
Cheers,
Kauy Gabriel.
10-09-2014 02:18 AM
Hi Kauy,
I didn't find any MAC_FLAPPING messages, but the bug fits the description. eth0 does have the mac address of the teamed NIC. I will try and test it later this month.
Thank you for your helping out.
Kind regards,
Ivo
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide