StoreVirtual 12.6 - Warning Event E010A0301:EID_NIC_ERRORS_EXCESSIVE

[Babeztar]

Hi folks,

I updated my storevirtual 4730 nodes to 12.6.

Since there I have NIC warnings on the primary node telling me:

Network Interface 'NICSlot2:Port1' error status = 'Excessive'

Network Interface 'NICSlot2:Port2' error status = 'Excessive'

The Network Interface 'NICSlot2:Port1' error status is 'Excessive' at  2.00%.

The Network Interface 'NICSlot2:Port2' error status is 'Excessive' at  0.93%.

What´s wrong or can I disable these message, cause two percentage or less I can accept?

Regards

Thomas

[Stor_Mort]

Hi Babeztar,

That means you have a network problem that results in significant packet loss. This will degrade storage system performance. 12.6 added a new check for this condition to provide visibility, because iSCSI SANs are much more sensitive to packet loss than other applications. Follow up with your network administrator to determine where the packet loss is occurring.

[Babeztar]

I checked with the networks admins.

There are no packet loss, just some pause frames, TX some more.

We raised a ticket and HPE supports tells to downgrade the firmware nic. Thats sounds spooky for me!

[Devform]

hi,
we have exactly the same problem and constellation

Version 12.60

error: the Network Interface 'NICSlot1:Port2' error status is 'Excessive' at  2.80%

we have also no packet loss, just some pause frames, TX

what is your solution?

you  downgrade the firmware nic.?
Thank you for a reply

[Stor_Mort]

If you do not think there is a problem on your network, open a support case with HPE so we can look at your logs in detail.

In almost all cases, we have found a problem with a cable or SFP, or a network problem like inadequate switch capacity or congestion. These are not storage issues and you may find it more efficient for your network support people to track it down rather than storage support.

[Devform]

We found the following solution for our problem:

We deactivated flow control (---> "flow control send off") on our switch interface aswell as the server.

 -- server type: 3430 FC ;   software version:12.6.00.0155.0;   Emulex-NIC  

This solves our problem:  The error disappears and there are no additional pause frames.

Thank you for all your advice, we hope this might be helpful for you aswell!

[BadHabit]

Hi Devform,

as I have the same issue, where excatly did you disable FlowControl, on Switch side, in the StoreVirtual itself and your hypervisor?

Thanks in advance

[mandrews]

I'm getting this issue on one of my 2 clusters,, 1 a 3x 4730 and the other 3x 4530, both configured identically in terms of NICs, flow control etc and both into identical switches configured the same, and on the 4730's 10GBe ports I am seeing 100% tx utilization on the switch constantly one one port on each of the nodes