My poor little 827-4v just started crashing.  I've been tweaking with a
Content Engine 507 (used as a proxy box, no WCCP support on the 827).  I'm
guessing that's what is hammering the box, but why is it the memory not
returning to free use?

falcon#sh ver
Cisco Internetwork Operating System Software
IOS (tm) C820 Software (C820-K2NOSV6Y6-M), Version 12.1(5)YB, EARLY
DEPLOYMENT RELEASE SOF
TWARE (fc1)
TAC:Home:SW:IOS:Specials for info
Copyright (c) 1986-2001 by cisco Systems, Inc.
Compiled Thu 08-Feb-01 01:00 by detang
Image text-base: 0x80013170, data-base: 0x80997DE4

ROM: System Bootstrap, Version 12.1(1r)XB1, RELEASE SOFTWARE (fc1)
ROM: C820 Software (C820-K2NOSV6Y6-M), Version 12.1(5)YB, EARLY DEPLOYMENT
RELEASE SOFTWAR
E (fc1)

falcon uptime is 5 minutes
System returned to ROM by reload at 18:32:58 PDT Fri May 25 2001
System restarted at 18:34:23 PDT Fri May 25 2001
System image file is "flash:c820-k2nosv6y6-mz.121-5.YB.bin"

CISCO C827-4V (MPC855T) processor (revision 0x502) with 23552K/1024K bytes
of memory.
Processor board ID JAD04390QBC (72633648), with hardware revision 0000
CPU rev number 5
Bridging software.
4 POTS Ports
1 Ethernet/IEEE 802.3 interface(s)
1 ATM network interface(s)
128K bytes of non-volatile configuration memory.
8192K bytes of processor board System flash (Read/Write)

Configuration register is 0x2

falcon#sh mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     6068088     1663920     1662060
1663268
      I/O    1700000     1048644      692208      356436      356436
355860

If I fire up my IE5 Offline Syncronize (Cisco Field Notices, Cisco
Headlines, Slashdot.org), within a few minutes it's dropped to:
falcon#sh mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     6362292     1369716     1350056
1345636
      I/O    1700000     1048644      692208      356436      356436
355860

This time it appears to have settled (probably because the CE507 has a large
portion of the content cached now).

After about 5 minutes:
falcon#sh mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     6693236     1038772     1038444
1029716
      I/O    1700000     1048644      692208      356436      356436
355860

Can we see a trend?  All the while, my Offline Syncronization is busy
pulling down sites (no doubt getting them faster with the CE507 reading
ahead and pulling down images to proxy to me before my PC even asks for
them).

Ok, so my question is, why is it getting hammered so hard?  I am running
CBAC Firewall and IPSEC with 6 tunnels (nothing going on with VPN), so that
all adds up with ACLs that packets have to go through.  But the overall
traffic speed is nothing staggering (less than 21K), but it's small web
traffic, so the pps are probably the more intense item.

Oh, about 10 minutes into it, and we're destined for memory allocation
problems as soon as we bottom out (which means telnet and console access
dies):
falcon#sh mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     6946532      785476      745176
735040
      I/O    1700000     1048644      692208      356436      356436
355860

I'm also curious why my Processor + I/O total doesn't come anywhere near
24mb (only about a third).

falcon#sh mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     7092000      640008      311376
357996
      I/O    1700000     1048644      692208      356436      356436
355860

At least the box is recovering now.  Before, the free and lowest would track
the same all the way 'till I lost my session when it bottomed out.  Now, it
at least appears to be recovering (It looks like the Cisco stuff finised,
and just Slashdot.org now and it seems to be crawling).

Anyway, any ideas on why this is happening so drastically?  The only time
I've ever killed ram is with routing protocols and tables getting to big (on
old 2500s with only 2mb).

Ok, and now it's really screwy.  I do a "show run" and it's blank.  I check
show version, and sure enough I've got something odd set (0x2).

Fixed that and bounced the box just before it was going to kill my telnet:

Configuration register is 0x2 (will be 0x2102 at next reload)

falcon#sh mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     7371692      360316      311376
348060
      I/O    1700000     1048644      692208      356436      356436
355860



falcon#show mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     7713780       18228       16028
776
      I/O    1700000     1048644      692208      356436      356436
355860


Heh, well, same results.  Within 10 minutes of the reboot and starting the
Offline Sync I'm back down and falling fast:
falcon#sh mem
                Head    Total(b)     Used(b)     Free(b)   Lowest(b)
Largest(b)
Processor   80FA04D8     7732008     7076268      655740      305376
370988
      I/O    1700000     1048644      694056      354588      354588
354012

Thoughts?  I'm going to go play with PIX OS 6 and confirm if the VPN Client
v3 requires the PIX to have 3DES.

--
Jason Roysdon, CCNP+Security/CCDP, MCSE, CNA, Network+, A+
List email: [EMAIL PROTECTED]
Homepage: http://jason.artoo.net/




Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=6004&t=6004
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]

Reply via email to