TCC2P MEM-LOW

Unanswered Question
Aug 27th, 2008

Why would TCC2P cards running 8.0.0 not have enough memory? We have a dozen or so circuits on a standard ring design. The Node reporting the alarm is a repeat site only. I haven't contacted TAC but I suspect they'll suggest a swact on the reporting shelf...

Thanks.

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Tom Randstrom Thu, 08/28/2008 - 12:06

This would not seem to be standard operation.

Per the Troubleshooting Guide:

MEM-LOW

Default Severity: Minor (MN), Non-Service-Affecting (NSA)

Logical Object: EQPT

The Free Memory of Card Almost Gone alarm occurs when data generated by software operations is close to exceeding the memory capacity of the TCC2/TCC2P card. The alarm clears when additional memory becomes available. If additional memory is not made available and the memory capacity of the card is exceeded, CTC ceases to function.

The alarm does not require user intervention. If the alarm does not clear, log into the Technical Support Website at http://www.cisco.com/techsupport for more information or call Cisco TAC (1 800 553-2447).

d.keeley Thu, 08/28/2008 - 12:28

Ya, read the same thing. It doesn't say much other than the obvious. Guess I'm hoping someone has experience with the root of the problem.

Thanks

tgimmel Sun, 08/31/2008 - 11:55

Do you have a large network or somehow have let it learn lots of routes? When the TCC2 get about 600 routes are so it hits the 50% memory alarm.

Tim

d.keeley Wed, 09/03/2008 - 05:07

This DWDM ring has 14 NE's, add another 6 or so for transponder shelves. The routing table on the reporting shelf currently shows 23 routes but the alarm has cleared. I'll check the routing table next time the alarm appears. Thanks for the info.

tsim496 Wed, 09/10/2008 - 17:30

I've found that if you're monitoring the chassis for performance statistics via SNMP, there tends to be a memory leak/exhaustion issue.

This can be followed by free running sync and TCC resets. In an extreme instance, I have seen a two side 6 node multishelf MSTP network reset several times per month. The only resolution was an upgrade to 8.51, which I was told contained enhanced memory management, and the omission of placeholders/support for older line cards.

d.keeley Thu, 09/11/2008 - 04:46

Well we don't monitor via SNMP as well we don't time any of the MSTP boxes which should rule out the free run sync issue. I don't believe I've seen TCC resets as a result of the MEM-LOW issue, other than those I've done myself to clear the alarm.

As for the upgrade to 8.5.1, as a result of the 2 degree 6 node multishelf resets, that's a direction I'm about to take.

Thanks for the info, very helpful

Actions

This Discussion