[Iwg-arbitration-committee] Official arbitration request for Mellanox MCX353A-FCBT and MCX354A-FCBT

[Rupert Dance]

Hi All,

 

Some early notification of good news. We updated the cluster to OFED 3.5 RC1
and the problem between the FDR cards and uDAPL has been resolved. The logo
is based on OFED 1.5.4.1 but we were unable to determine the root cause in
the uDAPL package despite checking with the uDAPL maintainers (one of whom
(Arlin) is on sabbatical). So based on our diagnostic results from OFED 3.5
RC1 we are going to pass the FDR cards with a comment that notes the failure
of uDAPL as packaged in OFED 1.5.4.1 and note that the problem can be
resolved by either updating to OFED 3.5 or by updating the individual
package. 

 

Thanks and please let me know if you have any problems with this solution.

 

Rupert

 

From: Yaroslav Pekelis [mailto:slava at mellanox.com] 
Sent: Thursday, July 19, 2012 8:50 AM
To: Iwg-arbitration-committee at openfabrics.org; Rupert Dance
(rsdance at soft-forge.com)
Cc: Hal Rosenstock; Amit Krig
Subject: Official arbitration request for Mellanox MCX353A-FCBT and
MCX354A-FCBT

 

Hi, 

 

Mellanox would like to fill official arbitration request for the following
devices:

 

MCX353A-FCBT and MCX354A-FCBT

 

The uDapl problem was not observed during the debug event and presented
itself only during the Validation event. 

 

More than that the test suite passes properly when working at QDR speed
which leads us to suspicious that test itself does not work properly with
Spec1.3 extended speeds such as FDR.

 

We will continue to work with UNH team in order to get to the root cause of
the reported failure. 

 

Kind regards, 

Slava

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/iwg-arbitration-committee/attachments/20120809/5ccb030a/attachment.html>