The Cisco Catalyst 6500 Just Ran

In my networking career, I worked with several Cisco Catalyst 6500 chassis switches. The first I remember was a 6509 running in a sort of hybrid mode. The Layer 2 switching engine ran CatOS with its set-based CLI. That was where VLANs, trunk ports, spanning tree, and related features were configured. The Layer 3 routing engine ran IOS. That was where routing functions were configured. The two engines passed data however they did in a way that I didn’t understand all that well as a somewhat young network engineer.

Sometime during my time administering that box, Cisco made it possible to convert that dual personality into a single persona running IOS. There was a process to it, well-documented by Cisco. I performed (or at least was part of, hard to recall) the upgrade to that 6509, and was a happier engineer when I no longer had to remember two styles of CLI to make the switch do what I needed it to do.

This 6509 was the centerpiece of a sprawling multi-campus network. It just ran.

At a different company, I inherited two Catalyst 6513s with what were, at the time, the big supervisor engines—the 720s. These were the core of a simply laid-out data center, feeding Catalyst 3750 stacks at top of rack or end of row. These 6513s also were the junction point for several other important networks we had at this shop, segmented in particular ways to meet security needs and regulatory requirements.

These 6513s were everything for this particular company. They just ran.

At yet another company, I found myself in charge of two Catalyst 4507Rs acting as the core switches of a growing little data center moving around an increasing amount of traffic. One of the 4507R line cards had an issue where it would log sporadically dropping packets due to a hardware problem. The 4507Rs weren’t covered with SmartNet and the network budget was tight.

While hunting around for a cheap solution to the problem, I found two 6509s sitting in a back room on a pallet. I inquired after them, and was told the previous networking tech lead took them out of service for whatever his reasons were. I fired them up, and found they were running fine. I could have asked for beefier line cards and sup engines, but after reading through the specs, I knew they could do the job despite some oversubscription.

One long Friday night and Saturday morning, we swapped out the 4507Rs for the 6509s, and the various little problems we’d had due to the failing line card went away. Until the day we took them out of service in favor of shiner Nexus gear, the 6509s just ran.

I’m not the only one that lived this “they just ran” experience.

Patrick wrote in to say…

“I read your email about Catalyst 6500 25th anniversary. I waited until today to send this so it could tick over to 17-year uptime. It’s powering a small computer room with a couple of racks of gear. Some might say it is reckless to continue to operate an older platform. It’s not internet-facing so big woop, we will keep it running even if we replace it just to see how long it can go for :-).”

Cisco Internetwork Operating System Software
IOS (tm) s3223_rp Software (s3223_rp-IPBASEK9_WAN-M), Version 12.2(18)SXF7, RELEASE SOFTWARE (fc1)
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 1986-2006 by cisco Systems, Inc.
Compiled Thu 23-Nov-06 03:02 by kellythw
Image text-base: 0x40101040, data-base: 0x42D30000
ROM: System Bootstrap, Version 12.2(17r)SX3, RELEASE SOFTWARE (fc1)
BOOTLDR: s3223_rp Software (s3223_rp-IPBASEK9_WAN-M), Version 12.2(18)SXF7, RELEASE SOFTWARE (fc1)
mr-sw uptime is 17 years, 10 hours, 32 minutes
Time since mr-sw switched to active is 17 years, 12 hours, 38 minutes
System returned to ROM by power cycle (SP by power on)
System restarted at 05:40:45 EDT Thu Apr 12 2007
System image file is "sup-bootdisk:s3223-ipbasek9_wan-mz.122-18.SXF7.bin"

Another reader says…

“Loving the newsletter, thanks for the notice of the Cat6k’s anniversary. A couple of years back my team inherited stewardship of 4 6500s with over 14 years of uptime on each about 18 months ago. Since then two have been replaced, and one did have a power failure, leaving one I’m trying to keep going until the campus it sits on completes its switching refresh. (I’m staying entirely anonymous to not give the game away to our infosec team). I’m probably sure you’ll get someone messaging in with a greater uptime, but this one stands at 15 years 5 months and 8 days so far 🙂 The problem is that it entirely implies that no-one upgraded any of them since they were put into production in 2008…”

Jeff shared…

“We have a 6504, that used to be our core switch, but has been out of commission for a few years. However, I’ve kept it powered on with only a management connection because I want to see how long it will stay running.”

CORE-6504#sh ver | i uptime
CORE-6504 uptime is 15 years, 50 weeks, 3 days, 16 hours, 38 minutes
CORE-6504#
CORE-6504#sh ver | i Version
IOS (tm) s3223_rp Software (s3223_rp-IPSERVICESK9_WAN-M), Version 12.2(18)SXF13, RELEASE SOFTWARE (fc1)
ROM: System Bootstrap, Version 12.2(17r)SX3, RELEASE SOFTWARE (fc1)
BOOTLDR: s3223_rp Software (s3223_rp-IPSERVICESK9_WAN-M), Version 12.2(18)SXF13, RELEASE SOFTWARE (fc1)
X.25 software, Version 3.0.0.
CORE-6504#
CORE-6504#sh ver | i revision
cisco WS-C6504-E (R7000) processor (revision 2.0) with 458752K/65536K bytes of memory.”

In memoriam

Sometimes I wonder why Cisco maintains the dominance in the networking space they’ve enjoyed for so long now. Their products are expensive. Technical support isn’t what it used to be. Their training programs, once (and perhaps still) the gold standard for networking education, are more Cisco-oriented now. The user community is frustrated by the diversity of operating systems, complexity of licensing schemes, and bugginess found across the mammoth product set.

The Cisco of 2024 isn’t the Cisco that made the Catalyst 6500. But we remember. And many of us believe that this purchase, this gear, this time… we’ll have the 6500 back again.

The new thing will just run.