[c-nsp] TestFabricCh0Health chassis or supervisor?

[Peter Rathlev]

Hi all,

We have what seems to be a slot in a WS-C6509-V-E chassis that acts up.
There is a WS-X6708-10GE module sitting in the slot and it has crashed
several times with the following log message:

%CONST_DIAG-SP-2-HM_LC_CRSH: Module 7 crashed due to unrecoverable errors, Reason: Failed TestFabricCh0Health

We tried reseating the card to no avail. The card was replaced after a
few crashes in a couple of months, but the new card is also crashing.
The old was HW revision 1.6, the new one i rev 2.2. In the same chassis
we have two other 6708 cards (HW revs 1.3 and 1.6) that are behaving.
The dysfunctional slot is 7 and the chassis is equipped thus:

 Slot  Module          HW-rev
 ----  --------------  ------
 1     WS-X6748-GE-TX  4.1
 2     WS-X6748-GE-TX  4.1
 4     WS-X6724-SFP    2.5
 5     WS-SUP720-3BXL  5.2
 7     WS-X6708-10GE   2.2
 8     WS-X6708-10GE   1.3
 9     WS-X6708-10GE   1.6

We are planning on replacing the current Sup720 with a Sup2T in the near
future, but are thinking that the error might also be chassis related.
Replacing the chassis would be easier to do at the same time as we are
replacing the supervisor and the three WS-X6708-10GE cards that cannot
work in Sup2T, but we would like to avoid the chassis change if it
doesn't do any good.

So does anybody know if "TestFabricCh0Health" can be strictly said to
relate to the supervisor or if the chassis could also be the culprit?

Thanks in advance.

-- 
Peter