[168655] in North American Network Operators' Group
RE: Twinax trivia check (was Re: Is there such a thing as a
daemon@ATHENA.MIT.EDU (Murphy-Olson, Daniel E.)
Sun Feb 2 17:49:36 2014
From: "Murphy-Olson, Daniel E." <dolson@mcs.anl.gov>
To: Jeff Kell <jeff-kell@utc.edu>, Bryan Tong <contact@nullivex.com>, "Jay
Ashworth" <jra@baylink.com>
Date: Sun, 2 Feb 2014 22:49:23 +0000
In-Reply-To: <52EEB58B.3060107@utc.edu>
Cc: NANOG <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
Most of the switch vendors have an "official" compatibility list, but I've =
found that generally the most common compatibility issue is active vs passi=
ve twinax. =0A=
=0A=
Brocade edge switches and nics are normally active only, which seems to com=
e up a lot - because most short cables are passive unless they are brocade =
branded. >5m is normally the cutoff for passive twinax. Pretty much every=
thing else I've encountered supports passive.=0A=
=0A=
For a while, the intel x520 nics, which are very common, didn't support act=
ive connections - but they have since released firmware that fixes this pro=
blem. =0A=
Netapp's lower end gear doesn't support active twinax. =0A=
=0A=
=0A=
________________________________________=0A=
From: Jeff Kell [jeff-kell@utc.edu]=0A=
Sent: Sunday, February 02, 2014 3:15 PM=0A=
To: Bryan Tong; Jay Ashworth=0A=
Cc: NANOG=0A=
Subject: Re: Twinax trivia check (was Re: Is there such a thing as a 10GBas=
e-T SFP+ transciever)=0A=
=0A=
On 2/2/2014 4:03 PM, Bryan Tong wrote:=0A=
> These cables are most commonly known as "Direct Attach Copper SFP+"=0A=
=0A=
The big issue appears to be that these are not always "consistently=0A=
functional" crossing vendor lines (sometimes product lines within the=0A=
same vendor). There does not appear to be any standardization in=0A=
place. Not sure how much of this is picky vendor software looking for=0A=
"branded" marks in their transceivers (e.g., Cisco "service=0A=
unsupported-transceiver") versus true incompatibilities.=0A=
=0A=
We have had issues in test cases crossing vendor lines (Cisco / Brocade=0A=
/ Dell / HP) with a "twinax" link that just simply won't work. If=0A=
anyone has a clear explanation or better understanding, I'm all ears.=0A=
Personal experience comes from only a few testbed cases.=0A=
=0A=
Jeff=0A=
=0A=
=0A=
