CS1K:6.0: Unable to manually swap CPUs in CS1000 HA Rel 6.0

[ktse1210]

Hi there,
Sigh from one issue to another.

Anyway, under LD 135, I tried carrying out "scpu" to manually switch cores but it fails to do so. The following is the result:

.scpu
OK
.
SRPT018 LCS: Graceful switchover not executed - local ELAN health is better than
remote.

CCED000




Here's the "stat cpu" and "stat health" print out:

.stat cpu

cp 0 1 PASS -- ENBL

SYSTEM STATE = REDUNDANT
DISK STATE = REDUNDANT
HEALTH = 14
VERSION = Jun 2 2009, 17:33:20
Side = 0, DRAM SIZE = 1006 MBytes

CP[0] located at IPMG [8 0 1]


cp 1 1 PASS -- STDBY

SYSTEM STATE = REDUNDANT
DISK STATE = REDUNDANT
HEALTH = 14
VERSION = Jun 2 2009, 17:33:20
Side = 1, DRAM SIZE = 1006 MBytes

CP[1] located at IPMG [0 1 1]

=========================================
Summary of Local System Resource (side 0)
=========================================
File Descriptors
-----------------
alloc 94
free 1954
total 2048

Unprotected Heap (bytes)
------------------------
alloc 74956536
free 463268056
total 538224592

Protected Heap (bytes)
----------------------
alloc 696572
free 3497732
total 4194304


.stat health

Local (Side 0, Active, Redundant):
Components without TIER 1 Health contribution:
==================================================
cp 0 1:In Service
ipb 0 1:In Service

TIER 1 Health Count Breakdown:
=====================================
sio8 0 1 1: 0002
sio8 0 1 2: 0002
fmd 0 1 1: 0008
eth 0 1 0: 0002
Local TIER 1 Health Total: 14


TIER 2 Health Count Breakdown:
=======================================
ELAN 16 IP : 172.17.116.28 Health = 2

Local AML over ELAN Total Health:2
Local Total IPL Health = 14

IPL connection history:7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7

Local TIER 2 Health Total:16

Remote (Side 1, Inactive, Redundant):
Components without TIER 1 Health contribution:
==================================================
cp 1 1:In Service
ipb 1 1:In Service

TIER 1 Health Count Breakdown:
=====================================
sio8 1 1 1: 0002
sio8 1 1 2: 0002
fmd 1 1 1: 0008
eth 1 1 0: 0002
Remote TIER 1 Health Total: 14


TIER 2 Health Count Breakdown:
=======================================
ELAN 16 IP : 172.17.116.28 Health = 0

Local AML over ELAN Total Health:0
Remote Total IPL health = 0

Remote TIER 2 Health Total:0


As you can see from the above, both the Local Core (Core 0) and the Remote Core (Core 1) have the same Health readings.

Both standalone CPPM CS's BIOS have been updated to version 18 (for Rel 6.0) and also went through a cold restart.

Switching of cores will only happen successfully if I reset the CORE 0 CPPM CS or physically take out the CPPM CS card.
However, when CORE 0 CPPM CS is back to operational mode, I tried to carry out the "scpu" command and it will fail as above.


Any ideas anyone?

 

[yyrkroon]

Both cores connected to the ELAN, doesn't look like it base on this info

TIER 2 Health Count Breakdown:
=======================================
ELAN 16 IP : 172.17.116.28 Health = 0

Local AML over ELAN Total Health:0
Remote Total IPL health = 0

Remote TIER 2 Health Total:0


the error message even say local elan health is better then remote and your print out proves it.

 

[ktse1210]

So, what's the remedial step?

Tell me if I have physically connected it correctly?

The 2x CPPM CS is connected to a Layer 2 switch via it's ELAN port and also connected directly to each other via a crossover cable on the HSP port.

Is this where it has gone wrong?

 

[ktse1210]

It's okay. I found out what the issue was. Sigh. For dual cpu's...I forgot to configure the inactive secondary elnk.
My bad.