Strange problem with AS5300

jimb · ‎09-03-2003

I have an AS5300 that, during hours of peak usage, drops all users off of the box. Access to the box via telnet is no longer possibe. VIA console connection, the box appears to be fine, but even though no users are displayed as connected to the box, none of the lines have cleared, and users dialing in get a fast busy. The box has not been reconfigured in any way, and has been reliable for 3 years. Anybody have an idea as to what this might be?

Results of show version list below.

murph#sh ver

Cisco Internetwork Operating System Software

IOS (tm) 5300 Software (C5300-IS-M), Version 12.1(5)T4, RELEASE SOFTWARE (fc1)

TAC Support: http://www.cisco.com/cgi-bin/ibld/view.pl?i=support

Compiled Thu 01-Feb-01 22:26 by ccai

Image text-base: 0x60008958, data-base: 0x60EA6000

ROM: System Bootstrap, Version 12.0(2)XD1, EARLY DEPLOYMENT RELEASE SOFTWARE (fc

1)

BOOTFLASH: 5300 Software (C5300-BOOT-M), Version 12.0(4)T1, RELEASE SOFTWARE (f

c1)

murph uptime is 11 hours, 23 minutes

System returned to ROM by reload at 09:42:29 UTC Sun Jan 2 2000

System image file is "flash:c5300-is-mz.121-5.t4.bin"

cisco AS5300 (R4K) processor (revision A.32) with 65536K/16384K bytes of memory.

Processor board ID 21672917

R4700 CPU at 150Mhz, Implementation 33, Rev 1.0, 512KB L2 Cache

Bridging software.

X.25 software, Version 3.0.0.

Primary Rate ISDN software, Version 1.1.

Backplane revision 2

Manufacture Cookie Info:

EEPROM Type 0x0001, EEPROM Version 0x01, Board ID 0x30,

Board Hardware Version 3.1, Item Number 800-2544-03,

Board Revision D0, Serial Number 21672917,

PLD/ISP Version 0.0, Manufacture Date 9-Aug-2000.

1 Ethernet/IEEE 802.3 interface(s)

1 FastEthernet/IEEE 802.3 interface(s)

196 Serial network interface(s)

192 terminal line(s)

8 Channelized T1/PRI port(s)

128K bytes of non-volatile configuration memory.

16384K bytes of processor board System flash (Read/Write)

8192K bytes of processor board Boot flash (Read/Write)

Configuration register is 0x2102

jimb · ‎09-03-2003

Update... the box crashed again. We retrieved the following messages from our syslog.

Local6.Critical %SYS-2-MALL0CFAIL: Memory alocation of 6000 bytes failed from 0x6037E22C, pool Processor, alignment 0

Local6.Critical 7802: Process="TTY Background", ipl=0, pid=23

Local6.Critical 7803: -Traceback=60356A40 60357EDO 6037E234 60379438 6036C7E4 60860E68 602C0EE8 602C1A38 602C1DB4 602DC4F0 6034D394 6034D380

Local6.Critical SYS-2 CFORKMEM:Process creation of Modem Autoconfigure failed (no memory)

vjokhoo · ‎09-03-2003

Do you see this condition when all 192 dialin ports are used simultaneously? I have a problem with the AS5400 with 384 ports where the ports remain active after the users disconnect. This happens only when all 384 ports are in use. It does not happen in the RAS where I have more modems that trunk lines. This has happened as recently as ver 12.3(1a)

jimb · ‎09-03-2003

Thought about that. The box terminates 8 PRIS, all in the same hunt group, so we disconnected one and terminated it on another box. The box seized up again, so we don't think it is necessaraly a load related issue.

From the reading I have done here on Cisco's site, it appears it might be a memory leak, but we can't be certain what causes it. Cisco advises monitoring free memory, and to watch to see which process might be failing to relase memory space, but that is tuff to do with customers getting knocked offline, so we terminated all PRIs on a backup NAS. At this point, if it is a memory leak, it appears to involve a process associated with calls.