CS1000 Media gateway unexpected reboot

[cinanza]

Hello All,
yesterday I have on my cs1000 5.5 an unexpected reboot of the
three MGC.

This is the history file with reason code of INI:

SYSDB RLS/ISS: 550J DATE/TIME: 21/4/2009 16:28:23 RECS: 347 SEQNO: 3297
% SYSLOAD ISSUE 5.50J: IDLE_SET_DISPLAY Âòáããï
% DSET000 DOWN LOAD 0 0 0 0 0 0 0 23798
%
% INI002 00000085
%
% SRPT752 INI 0: INI completed in 5 seconds
%
% SRPT179 STARTUP: IPMG Controller task completed successfully


That INI002 is "I/O device fault, where xx is to the maintenance display code in.", but where can I find the
maintenance display code of 000000085 ?


In each MGC log I've found this reasons of reboot:


0172] 21/04/2009 17:34:07 SRPT0016 OMM: IP link is DOWN between Primary Call Server and IPMG[0 0]




[0269] 21/04/2009 17:38:07 SRPT0107 Hardware reset reason = Hard Reset [register value = 0x0]



Any ideas ?

Thanks in advance.

Cinanza

 

[Tman45]

Well ... you will want to check the IPMG patch level as I am suspecting you are less than a level of CAM03. I am trying to remember if we still had unexpected reboots happen on AH19 loadware ... please post your issp and mdp.

Another thing to look at is that I didn't think you are supposed to use IPMG 0 0 location based on the way Nortel resources are setup. You should start at IPMG 4 0.

 

[allenmac]

It is not a problem using IPMG 0 0.

I think this is a misunderstanding that comes from the CS1000s upgrade to CS1000e procedure where the call server ends up in IPMG 0 1. This is to compensate for the "empty" call server cabinet in the 1000s during the database conversion.

I have new CS1000e installations where I use IPMG 0 0, and also a CS1000s upgrade to CS1000e HA where the 2nd core is in IPMG 0 0, and I don't have any problems with these systems.

 

[Tman45]

Ah, so guess that means our vendor could have started at IPMG 0 0 ... oh well, they started at 4 0 and it must be from the 'misunderstanding' you are talking about. Thanks for setting me straight allenmac

So ... cinanza ... do you have a MGC loadware of CAM03 or above? Lastest and greatest (and on the new DepList) is CAM04. I do believe we still had unexpected reboots on AH19 loadware.

Another thing to consider is your SRPT 107 (hardware reset) - it is not listed in the Nortel docs. I had a problem arise that was not listed either and it ended up being a system issue (secondary CPPM was trying to take clocking away from the primary CPPM - fixed in CAM03)

SRPT0106 LCS: Join denied. Incompatible IRQ order between both sides.
Action: Contact technical support.
Severity: Major. Critical to Monitor: No. SNMP trap: Yes
SRPT0108 VO login Password retry failed 3 times. Tried to login as TN: 0x%x on the set
TN: 0x%x.
Action: Let the login lock expire in one hour, or disable/enable the set to login
through LD 32.

 

[MagnaRGP]

The ONLY downside to using IPMG in loop 0 0 is that you will lose 1 dsp on the 32 port DB on that MGC. You can't have a DSP in TN 0 0 0 0.

There were issues early on with CC6 not liking 0 0 X X TNs but they seem to have gone away.

I try to start at 4 0 if building from scratch, but any DB conversion will start with 0 0.

If you have a CS1000E with CPPIV's you can put your CS's in 0 0 and 0 1 then start your TN's with 4 0.