IBM E10, Service Manual, Page: 255 / 279

Appendix C. Isolating a CPI Diagnostic Progress Code Stop
MAP 4030: Isolating CPI Diagnostic Progress Code Stop
Attention: This is not a stand-alone procedure.
Customer disruption may occur if microcode and power boundaries are not in the proper conditions for this
service action. Ensure that you start all service activities in Entry MAP for All Service Action in chapter 2 of
the Enterprise Storage Server Service Guide, Volume 1.
Attention: The FRUs and cables in this procedure are ESD-sensitive. Always wear an ESD wrist strap
during this isolation procedure. Follow the ESD procedures in
″
Working with ESD-Sensitive Parts
″
in
chapter 3 of the Enterprise Storage Server Service Guide, Volume 2.
Description
Normally, CPI errors detected by the CPI diagnostics during the cluster bay power on will fence that CPI
interface and create a problem requesting service. It is possible for cluster bay to stop with the CPI
diagnostics displaying a 10 digit progress code in the cluster bay operator panel. This MAP uses that code
to determine the possible failing FRU.
Procedure
1. The cluster bay stopped with a 10 digit CPI diagnostic progress code displayed in the cluster bay
operator panel. This code contains information that will be used with the code listing below to
determine the failing FRUs.
2. This step contains an example of how to convert the progress code to the probable failing FRU(s) and
locations.
v
Assume that cluster bay 1 stopped with SD010000156, the first entry in the table below.
v
The last digit of the progress code
″
6
″
, is the CPI interface that was being tested when the cluster
bay stopped.
v
There are four FRUs listed. The first column is the Resource Name of the FRU, while the second
column is the FRU Name. The last column is the probability in % that the FRU will fix the problem.
v
At the top of the table, next to
″
Cluster Bay with progress code -->
″
is a group of columns labeled
″
Cluster Bay 1
″
and another group labeled
″
Cluster Bay 2
″
. Refer to the
″
Cluster Bay 2
″
group.
v
Beneath that is
″
Last digit of progress code -->
″
. In this case it is a
″
6
″
.
v
Go directly down and notice that the number in the row for each FRU lists the slot or bay for the
FRU.
v
If you choose to replace the IOA Card (70% probability to fix the problem), it is in I/O Planar slot 9
of cluster bay 2.
v
The I/O Planar is in the failing cluster bay.
v
FRU resource name for the I/O Attachment Card is rsioaxy, where x is the cluster bay the IA resides
in, and y is the Bay the IA connects to.
v
FRU resource name for Bays are rsiomy, where y is the numeric value representing the Bay in error.
3. To replace the FRUs, ensure the customer is not using the other cluster bay. Power off the 2105 Model
E10/E20 and replace the FRU(s). Power on the 2105 Model E10/E20.
v
If both cluster bays display READY, the problem is fixed. Go to
″
MAP 1500: End Service Action
″
in
chapter 3 of the Enterprise Storage Server Service Guide, Volume 1.
v
If the cluster bay still stops, replace the remaining FRUs, then call the next level of support.
© Copyright IBM Corp. 1999, 2000
239