[f-nsp] Incoming CRC-Errors on remot side but no outgoing errors on my switch but there are WriteDrops.

[Gunther Stammwitz]

Hello my dear colleagues,

I'm seeing some WriteDrops on the primary management module of my foundry
BigIron 4000. I have been searching for something since I a lot of
CRC-Errors on some Switches connected to my BigIron4000. The interesting
thing was that I couldn't see any outgoing errors but the guys on the other
end was incoming CRCs. The connection to the remote-guy is terminating on my
port 2/1 and I had already moved it to 2/8. 
Module 2 is the backup management module, module 1 the active one. Both are
B8GMRs (gen2).


The problem is happening with traffic on module 1 and module 2 but all of
the traffic is passing trough module 1.

I have enabled the "debug hw"-command and the log shows:

Jun 16 02:46:23:I:System: Slot 1 Write Sequence Drop 26312 within 5 minutes.

Jun 16 02:45:53:I:System: Slot 1 Write Sequence Drop 13851 within 5 minutes.

Jun 16 02:32:23:I:System: Slot 1 Write Sequence Drop 1 within 5 minutes. 


Bigiron4000>show backplane 
_______________________________________________________________________
Slot  Mod   FreeQ    DMADrop    BPDrop    WriteDrop     Last     
-----------------------------------------------------------------------
 1   B8GMR    898       0          0      47480     D:0  H:0 M:32S:3
 2   B8GMR    916       0          0          0          NEVER  
 3   B24E     900       0          0          0          NEVER  
As you can see a show backplane outputs a lot of WriteDrops.



We had been experiencing some problems with this switch a week ago: the
(unused but present) B24E-module in slot 4 was shown as being removed
although no one touched it. Removing and inserting the module didn't help so
I guess it most have died. The module is still inserted in the switch but I
issued a "no module 4"-command and it is no longer being shown. I have now
removed the broken module and cold-powered the switch but the WriteDrops and
the incoming crc errors on the remote side are still there.

Bigiron4000>Show backplane (after cold start and a few minutes of
operation.)
_______________________________________________________________________
Slot  Mod   FreeQ    DMADrop    BPDrop    WriteDrop     Last     
-----------------------------------------------------------------------
 1   B8GMR    914       0          0       4491     D:0  H:0 M:7 S:4
 2   B8GMR    916       0          0          0          NEVER  
 3   B24E     900       0          0          0          NEVER  

Is there some sort of testing mode I can use to start some diagnostinc
procedures on the modules?


That's what the remote switch is seeing after a few minutes of operation
with a little bit of traffic.
Ethernet                     Packets               Collisions        Errors
Port              [Receive            Transmit]   [Recv  Txmit]  [InErr
OutErr]
2/1                  88626                 1620       0       0     156
0


Any idea what's wrong here?
Maybe my chassis / backplane is broken or is it the active managament module
in slot1?


Thanks,
Gunther