Hello,
Our servers connected to different Cisco WS-C2960G-24TC-L switches with
different cable length and located in different L2 segments and
rx_short_length_errors counting only with 82574L NICs.
Best regards,
Aleksey
From: Terry Hardie [mailto:[email protected]]
Sent: Friday, June 01, 2012 5:03 AM
To: Aleksey Chudov
Cc: [email protected]
Subject: RE: [E1000-devel] Intel 82574L rx_short_length_errors
I've tried 9 ways to Tuesday to capture these frames, but I think they're
dropped before they get to any sniffer, since they are invalid L
Let me ask you this:
Are you servers that are doing it connected to a Juniper switch by chance?
Terry
From: Terry Hardie [mailto:[email protected]]
Sent: Wednesday, May 30, 2012 12:54
To: 'Aleksey Chudov'
Cc: '[email protected]'
Subject: RE: [E1000-devel] Intel 82574L rx_short_length_errors
The problem is that is only happens on servers getting a reasonable (> ~5000
pps and greater than 50Mbit). So, idle servers' error counters do not
increase.
Nonetheless, I still tried a sniff on a server getting at least 1
rx_short_length_errors per second. I could not find any anomalies in the
sniff. I was specifically looking for short frames. Another interesting
thing is that the rx_short_length_errors stop counting up as soon as the
interface goes into promiscuous mode. Not sure what this means.
I currently have 27 ethernet ports behaving like this.
Terry
From: Aleksey Chudov [mailto:[email protected]]
Sent: Wednesday, May 30, 2012 9:53
To: 'Terry Hardie'
Subject: RE: [E1000-devel] Intel 82574L rx_short_length_errors
Hello Terry,
Thank you for your response. Unfortunately I was unable to fix the problem.
Looks like this NIC is counting some service packets, as errors.
Could you try to run tcpdump on one of the idle servers to check and see
which packets arrive at the same time as errors counting. Of course if you
have some spare time.
Please forward this message to a mailing list to be seen by developers.
Best regards,
Aleksey
From: [email protected] [mailto:[email protected]] On Behalf Of Terry Hardie
Sent: Tuesday, May 29, 2012 11:54 PM
To: [email protected]
Subject: Re: [E1000-devel] Intel 82574L rx_short_length_errors
Hi Aleksey,
I'm having exactly the same symptoms you describe on a few dozen servers
(rx_short_length_errors at a low rate). What interesting, is we have
hundreds of these servers all over the world, and I only have 1 site that is
doing it, and only the servers connected to Juniper EX4200. I have the exact
same servers at the same site connected on the same VLAN to a Cisco 6509,
and they're not getting the errors. ALL the servers on the Juniper switch
are getting the errors, and have been for around 4 days now. Prior to that,
the servers have been up for over 200 days with NO errors. VERY strange.
I've just subscribed to the e1000-devel list. Feel free to copy/respond to
this email on the list.
Terry
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
E1000-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel® Ethernet, visit
http://communities.intel.com/community/wired
