ROLM 9751 Missing Transaction Error

[dwmeis1]

I have the following error on a multi-node ROLM MOD 70:
-----------------------------------------------------------------------
-- DISK -- -- ACTIVE MEMORY --
STATUS STATUS SIDE

NODE 1: OK OK CPU1
NODE 2: OK - 1 CPU1

-----------------------------------------------------------------------

ERROR SUMMARY
=============

NODE 1: NO ERRORS

NODE 2: ACTIVE CPU LACKS DATABASE UPDATES
RECOMMENDATION: LOAD STBY FROM DISK,
SWITCHOVER, THEN LOAD STBY MEM


Can anyone tell me the commands to correct this and from which node to execute them in?

 

[keyset6]

The software on the standby and active sides of Node 2 don't match, likely because a LOAD STBY MEM ALL wasn't done after software changes. You need to update the standby side. I think the command is LOAD STBY (9005.6.84, or whatever the release # is). After a few minutes do LI STATU SYS S. This lists the standby side. Be sure before following to the next step that SYSTEM_STATUS = NORMAL and WDT = ENABLED.

This will cause the system to switchover and likely cut off calls. After verifying the above atatus I think the command is STOP WDT A. This stops the watchdog timer on the active side which forces a switchover to the standby side. There may be another way to do this... SWITCHOVER U for unconditional? Been a long time. Either way after switching into the standby side do START WDT S, this restarted the watchdog timer stopped previously on the former active side. Do LOAD STBY MEM ALL. The CPUs should then have the same config.

Please note it's been years since I've done this - so these notes are just from memory. You should get more advice and/or verification that this is the correct process before proceeding.

 

[kevin906]

CNFG_STATUS can display multiple messages about the condition of CNFG in each node.
As a first step always switch the node in question back to the other CPU.
If the STBY memory and CNFG matches that will solve the issue.
If it comes up missing transactions on both CPUs the node in question will need to be downloaded.
You do not need to download nodes that are not showing a problem.
To switch a node back over—SWITCH UNCON MINOR--does not disrupt calls.
This is a CLI command.

If both CPUs of the node in question are missing transactions and a download
isn’t practical due to the time of day a RESET CONSISTENCY can be done
to allow for moves and changes to be done in the interim. A download should be
done as soon as possible following a RESET CONSISTENCY command.
RESET CONSISTENCY is executed from the CNFG command line.
This command will not affect any telephony.

To recover if needed.
From the node with the issue
Load STBY 9005.X.XX where this is the software release of the system
You will eventually see the message about load stby succeeded.
Don't switch over until system status changes from FINIT to NORMAL
LIST STATUS SYS X S where X is the node number being reloaded.
When system status is normal --START WDT x S where X is the node number
SWITCH UNCON MAJOR to switch to the reloaded CPU. This disrupts all calls.
Disable any previously disabled tests, check for telephony, check CNFG_STATUS
LOAD STBY MEM to correct the standby CPU

 

[PERRYPJ]

I would try to switch Node 2 back to the other side and se if the error is still there It may save you a lot of time

OLD ROLMEN WORKING ON NORTELS AND AVAYA