Integrity Servers
cancel
Showing results for 
Search instead for 
Did you mean: 

superdome Machine Check Analyzer output

 
SOLVED
Go to solution
stephen peng
Valued Contributor

superdome Machine Check Analyzer output

dear all,
I've got a superdome HPMC, and the ts99 was analyzed by HP and part of the analysis:
Problem: CC5, MHG FE Err - (11)Recall to local PD times out.
Possible Cause 1: CC5, internal failure.
Possible Fix 1: Replace CC5.
Possible Cause 2: Destination CC5, failure.
Possible Fix 2: Replace destination CC5.
Possible Cause 3: It is possible that a 3rd party CC has caused the timeout.
Possible Fix 3: Look for other problems that indicate other CC suspects in the same PD.

Problem: (17)Cell 7, POUT2 signaled Uncorr Err: Rd or Flush time-out to
cell in PD.
Warning: The FRU identification for this error is not precise and may
be difficult. Look for the same cell number indicated in multiple
problems to indicate the most probable suspect.
Possible Cause 1: The destination cell (CC0) has blocked for too long.
Possible Fix 1: Replace cell (CC0).
Possible Cause 2: The source cell (CC0) was blocked for too long.
Possible Fix 2: Replace cell (CC0).
Possible Cause 3: An element in the packet path between the source
cell and destination cell blocked a packet for too long.
Corrective Action 3: Check for other errors listed for possible
source cell errors and perform the indicated fix(es).

Problem: (17)Cell 7, POUT3 signaled Uncorr Err: Rd or Flush time-out to
cell in PD.
Warning: The FRU identification for this error is not precise and may
be difficult. Look for the same cell number indicated in multiple
problems to indicate the most probable suspect.
Possible Cause 1: The destination cell (CC0) has blocked for too long.
Possible Fix 1: Replace cell (CC0).
Possible Cause 2: The source cell (CC0) was blocked for too long.
Possible Fix 2: Replace cell (CC0).
Possible Cause 3: An element in the packet path between the source
cell and destination cell blocked a packet for too long.
Corrective Action 3: Check for other errors listed for possible
source cell errors and perform the indicated fix(es).
can someone give some conclusion about it? I consider that I need to replace Cell, but which? Cell 0 or Cell 7? and what is keyword "CC5" for? cpu controller or cell controller?
by the way, where could I get this Machine Check Analyzer tools?


6 REPLIES 6
melvyn burnard
Honored Contributor

Re: superdome Machine Check Analyzer output

Well if HP decoded this for you, then they should also be making recommendations as to what should be changed, and get a hardware call logged, provided you have hardware over with hP.
The tools are not for public distribution as far as I recall.
My house is the bank's, my money the wife's, But my opinions belong to me, not HP!
Benoy Daniel
Trusted Contributor

Re: superdome Machine Check Analyzer output

Could you login to MP, select show logs and see the latest SEL (system event log).
P Muralidhar Kini
Honored Contributor

Re: superdome Machine Check Analyzer output

Hi Stephen,

>> I've got a superdome HPMC, and the ts99 was analyzed by HP and part
>> of the analysis:
If this was analyzed by HP, then what was their recommendation.
You might want to check with them as to what needs to be done in order
to solve the problem.

Check the following link -
http://www.hpux.co.kr/HW/server/Superdome/sms-1.2/release_notesv1_2.pdf
This talks about "SD High Priority Machine Check (HPMC) Analyzer"

Regards,
Murali
Let There Be Rock - AC/DC
stephen peng
Valued Contributor

Re: superdome Machine Check Analyzer output

I must admit that replacing cell would do the system good, what I want more was that what CC5 and POUTS stood for. you know, as system adminstrator, we would meet this situation, and we sometimes want to analyze by ourself to figure out what was really happening
P Muralidhar Kini
Honored Contributor
Solution

Re: superdome Machine Check Analyzer output

Hi Stephen,

>> what I want more was that what CC5 and POUTS stood for
Check the following link for more information -
http://www.dectrader.com/docs/set4/SDSYS-HW-INFO.pdf

CC Stands for Cell Controller
Refer Section - "9 Overview: Cell Controller (CC)"

POUT Stands for Outbound path.
Refer section - "How Does the Cell Controller Work?"

Hope this helps.

Regards,
Murali
Let There Be Rock - AC/DC
stephen peng
Valued Contributor

Re: superdome Machine Check Analyzer output

Murali,
Thank you! That was exactly that kind of answer I am looking for.