Conference proxy::laser

1264.0. "missing path to HSJ40 problem" by CSC32::I_WALDO () Tue May 06 1997 12:17

Got a strange problem, would appreciate any comments.

Config:

4 DEC7640 with three CIXCDs each.
Three star couplers
26 HSJ40s and 6 HSC95s
All CIXCDs have the same firmware, all HSJ40s have the same firmware, etc.


One of the DEC7640s is having a GOOD-to-BAD, BAD-to-GOOD cable problem 
with one of the HSJ40s.  This problem has lived thru at least three 
reboots of the DEC7640.  I just don't understand why only one node is 
having a problem with only one path on one HSJ40!  VTDPY on the HSJ 
shows data flowing on both paths.  Show cluster/continuous shows all is
normal except the one missing cable to that one HSJ from one DEC7640.

The path will stay good for about 10 seconds and then go bad again. 
Happens 50 times or there abouts and then stays bad for several hours 
before starting the cycle again.

Current plan is to replace the CIXCD that goes to the star coupler for
the HSJ40 with the missing path, theory being weak drives on the CIXCD.
Seems rather far fetched to me but I don't have any better idea!

******************************** ENTRY  543 ********************************


Logging OS                        1. OpenVMS
System Architecture               2. Alpha
OS version                           V6.1
Event sequence number             0.
Timestamp of occurrence              02-MAY-1997 04:50:03
Time since reboot                    0 Day(s) 0:00:23
Host name                            OHTVS2

System Model                         DEC 7000 Model 640

Entry type                      100. Logged Message


---- Device Profile ----
Unit                                 OHTVS2$PNB0
Product Name                         CIXCD (XMI to CI Adapter)

---- MSCP Logged Msg ----

Logged Message Type Code          3. Port Message

Error Type/SubType            x4100  Cable Status Change, Path #0. went from
                                     GOOD to BAD.
Count - Remaining Retries        50.

Error Count                       1.
Local Station Address     x000000000000000E
Local Station ID          x0000000000001AE8
Remote Station Address    x0000000000000009
Remote Station ID         x0000420010090320

NPPD Opcode                     x00  Port Message
NPPD Channel Index              x00
NPPD Flags                    x4001  Response Generated
                                     Path Select:   Path A
NPPD Status                   x0229  Command Failed.
                                     Status Type:   No Path
                                     Path A Status:   No Response
                                     Path B Status:   ACK (Success) or Not used
NPPD Body Opcode                x05  IDREQ
NPPD Body Flags                 x00
NPPD Destination Subnode          0.
NPPD Destination PGRP Prt         9.
NPPD Source Subnode               0.
NPPD Source PGRP Port             0.


******************************** ENTRY  669 ********************************


Logging OS                        1. OpenVMS
System Architecture               2. Alpha
OS version                           V6.1
Event sequence number         31006.
Timestamp of occurrence              02-MAY-1997 05:20:33
Time since reboot                    0 Day(s) 0:30:53
Host name                            OHTVS2

System Model                         DEC 7000 Model 640

Entry type                      100. Logged Message


---- Device Profile ----
Unit                                 OHTVS2$PNB0
Product Name                         CIXCD (XMI to CI Adapter)


---- MSCP Logged Msg ----

Logged Message Type Code          3. Port Message

Error Type/SubType            x4102  Cable Status Change, Path #0. went from
                                     BAD to GOOD.
Count - Remaining Retries        50.

Error Count                       2.
Local Station Address     x000000000000000E
Local Station ID          x0000000000001AE8
Remote Station Address    x0000000000000009
Remote Station ID         x0000420010090320

NPPD Opcode                     x01  Port Message Received
NPPD Channel Index              x00
NPPD Flags                    x0000
NPPD Status                   x0100  Packet Processed Correctly
                                     Status Type:   OK Status
                                     Receive Path:   Path A
NPPD Body Opcode                x0B  ID
NPPD Body Flags                 x10
NPPD Destination Subnode          0.
NPPD Destination PGRP Prt        14.
NPPD Source Subnode               0.
NPPD Source PGRP Port             9.

Perhaps bad cable(s)?

    This is a signal integrity problem that has a history of being hard to
    resolve. You are dealing with the combinations of output levels, losses
    due to the path and the gain and noise immunity of the input. I have
    seen one node report the path going from good to bad while the other
    node has no problem. One of the best ways to resolve is to move
    components around and look for the problem to move. often this cause
    the problem to disappear or become intermittent. Keep good notes.
    Remember that the components you are looking at are:
    	T2080 modules
    	XMI backplane slots
    	paddle cards
    	bulkhead cable assemblies
    	cables
    	SC008 star coupler.
    Make sure that all of the terminators on the star are on and tight.

    I could also be noise getting into the system. Make sure that the SC008
    panels are isolated from each other (ie. path A and path B not touching
    and not touching a ground). 

    Good luck

    Thanks for the list of things to check.  We will start eliminating
    items this weekend.  Was afraid there wasn't going to be a quick and
    certain fix.