Monday, March 15, 2010

Rant:“Investment Protection” on the Cisco C6500 — It Isn’t Quite That Simple

April 25, 2009 by Greg Ferro · 5 Comments 


Omar Sultan over at the Cisco Data Centre blog in this post claim­ing that the Cisco 6500 has a fant­astic invest­ment pro­tec­tion story.

Let’s face it, the Catalyst 6500 is the poster child for invest­ment pro­tec­tion. It has been ship­ping since 1999 and is still hap­pily chug­ging along.

Urrrmppffff.
Excuse me while I barf up my my breakfast. Sorry about this. 

Hang on, won't be a sec.


Right, where were we.

The real story isn’t quite so simple. While I agree that that the ori­ginal Catalyst 6000 (not 6500) was released in 1999 as a Layer 2 device, it has had more than half a dozen super­visors. The first Supervisor 1 wasn’t very good, Supervisor 1A (which fixed all the bugs) then the Supervisor II and IIA, Supervisor III (in sev­eral vari­ants) and then on the Supervisor 720 (in its many vari­ants) and, in back to the future blitz, the Supervisor 32 with a cut­down fea­ture set. Lets not for­get the Switch Fabric Modules (WS-​​C6500-​​SFM, WS-​​X6500-​​SFM2)

While the chassis has been con­stant (except for back­plane upgrade for Cat6000 and cer­tain mod­els amd early 6500’s around 20002001) you could make the claim that every Supervisor upgrade is a “rip and replace”.….if you were a mar­ket­ing per­son. After all, the chassis doesn’t get replaced.…. or does it ?

For those of us who actu­ally work in the field, repla­cing super­visors IS A RIP AND REPLACE pro­ced­ure. As each super­visor added fea­tures and drastic­ally changed the tech­nical land­scape, I rarely just “replaced the super­visor” as the risk was took great. We would always buy another chassis and “cutover” to the new unit.

Lets not for­get all the line cards — the 61xx cards using the 32GB shared bus, the 630xx POE, 65xx cards that con­nec­ted using single fab­ric lines and the 67xx cards that finally, ten years later meet the prom­ises that were made in 1999. Promises that seem con­veni­ently forgotten.

Oh yes, I remem­ber the Cisco Account Mgr telling us in tones of sepulchural joy about the amaz­ing fea­tures of the C6000 back­plane, and its future cap­ab­il­it­ies. How we purred over the spe­cific­a­tion and got excited. As the years passed, we for­got about the prom­ises and lived with the real­ity. (And they are also true of the Nexus 7000 as I wrote about early last year)

The beau­ti­ful future we were going to have, hold­ings hands as we ran through fields of beau­ti­ful flowers. couple_in_the_park-sml.jpg

OH YEAH, LIKE THAT HAPPENED !

The archi­tec­tural limitations

As someone who has spent over ten hours this week explain­ing the lim­it­a­tions of the C6500, and how little band­width the back­plane has and how con­strained this makes your designs. The con­stant review of all the design tricks you need to make it run the way you inten­ded, espe­cially under high load.

Only two fab­ric inputs per card, no load bal­an­cing of those inputs. The Fabric has head of line prob­lems under load, the short­com­ings of over­sub­scribed line cards. The Supervisor 32 with truly lim­ited per­form­ance. Lets not for­get these things. What about all those line cards, and the rules and tricks you need to know about their per­form­ance ? How many queues per port ? Per port group ? Which back­plane is it using — the shared bus or the fab­ric, how many fab­ric con­nec­tions does it have ?

Here’s to the Catalyst 6500

But any­way, here’s a toast to the Cisco Catalyst 6500. Its a teen­ager now with hair in its armpits and a gruff voice. For all its cranky ways, and tan­trums, at least it has worked most of the time. It hasn’t been a bril­liant product, or the best at any­thing. But for most of us, that’s what we want — some­thing that just works.

On the other hand, I can still get sup­port. Maybe that’s what “invest­ment pro­tec­tion” really means.

Note to Cisco /​ Omar: This means less mar­ket­ing and more con­cen­tra­tion on what really keeps Cisco in busi­ness — tech­nical sup­port. The IT man­ager is usu­ally happy when the engin­eers are not grumbling about the product. Not because of some rub­bish about “invest­ment protection”.

Rant over.

PS: Cisco makes a lot of money out of maintenance

A related asser­tion I see folks try­ing is that Cisco doesn’t innov­ate because their switches are X years old. I will tell you, that as a com­pany, we are quite pleased to see a data cen­ter filled with dusty Cisco chassis. It means we have given our cus­tom­ers the abil­ity to add cap­ab­il­ity and capa­city based on their needs. It means we respect the invest­ment cus­tom­ers have made with us and it means our engin­eers are smart enough to pro­tect that investment.

This really means “we made a shed load of money out of selling main­ten­ance on the same old kit ever year. We’ve been rak­ing it in!”

Please rate this post:

  Why Rate Posts?
1 Star - It\\\'s Crud2 Stars - It\\\'s Tosh3 Stars - Something\\\'s missing4 Stars - Needs works5 Stars - Good Enough6 Stars - Good7 Stars - Excellent8 Stars - Brilliant9 Stars - Astonishing10 Stars - Awesomely Godlike? (No Ratings Yet)
Loading ... Loading ...

Comments

5 Responses to “Rant:“Investment Protection” on the Cisco C6500 — It Isn’t Quite That Simple”
  1. Brad Hedlund says:

    Perhaps just a typo, but the Supervisor 32 is not Layer 2 only.

  2. Omar Sultan says:

    Hey Greg:

    Always appre­ci­ate a good rant. :)

    In this case, it also helps make my point. I did not say the Catalyst is a mono­lithic box that has been unchanged in 10 years. It has evolved over the years in, as you point out, an num­ber of incre­mental steps, and some­times the upgrades were not an incon­sequen­tial effort. And we have learned from that and come up with things like ISSU on the 6500 and a whole host of fea­tures on the N7K.

    My point, how­ever, is that the change is driven at the customer’s pace because they want or need new fea­tures need addi­tional capa­city, not because we got bored and EOL a chassis and decide to not sup­port it and pour our invest­ment into our cool new plat­form. The Catalyst gives cus­tom­ers gran­u­lar con­trol to only apply upgrades where they need – they can keep a con­sist­ent oper­a­tions and man­age­ment envir­on­ment without hav­ing to forgo the abil­ity to incor­por­ate new tech­no­lo­gies as needed, where they are needed. For example, f you need to deploy a chassis full of 10GbE is it prop­erly sim­pler to buy a new chassis already loaded? Probably, but I don’t find most cus­tomer deploy a new tech­no­logy that way. They will take a new tech­no­logy and deploy it incre­ment­ally and spar­ingly where it is cost-​​effective or func­tion­ally required, then then shift buy­ing pat­terns as the new tech­no­logy becomes main­stream. For the rest of the net­work gear, if its doing what the busi­ness needs it to do, leave it alone and let it get dusty.

    As for the post­script, a good num­ber of our cus­tom­ers buy their sup­port through chan­nel part­ners, self-​​spare and the like, so I don’t think we are “rak­ing in” all that you think we are. But again, it helps make the point. Customers can con­tinue to feel com­fort­able with their invest­ment because of con­tin­ued sup­port (heck, we just stopped sup­port­ing Catalyst 5xxx hard­ware last year – 12 years after we intro­duced them and 5 years after we stopped selling them). The lack of hard­ware churn means that cus­tom­ers can invest in their own spares to save money. The lack of plat­form churn also means they are not con­stantly wast­ing time dong test and qual on new sys­tems. The oper­a­tions and man­age­ment con­sist­ency means that cus­tom­ers don’t have to keep send­ing folks off to get retrained and the net­work folks them­selves have highly mar­ket­able skills.

    Finally, I do agree with you Greg, that, ulti­mately, we want the engin­eers happy with the product and we con­tinue to focus on that through both engin­eer­ing and TAC. But, I am also guess­ing, some­where up the line, whomever is pay­ing the bills is also fairly pleased that we worry about invest­ment pro­tec­tion too.

    Have a good weekend,

    Omar
    Cisco Systems

    • Greg Ferro says:

      1– you may not have said it’s mono­lithic, but it was cer­tainly implied in the art­icle.
      2 — Cisco main­ten­ance remains “reas­sur­ingly expens­ive”, so much so that it gives part­ners a chance to offer it cheaper. This has caused me prob­lems for years, since part­ners con­sist­ently do main­ten­ance badly. Bean coun­ters always choose the cheapest and are quick to com­plain when we can’t meet SLA’s. Of course, “I told you so” doesn’t work very well. (That’s a no win pos­i­tion there).
      3 — What you call “invest­ment pro­tec­tion” isn’t a proper defin­i­tion. But that is why I have a blog so I can com­plain and ima­gine that I have changed the world. Which, of course, I didn’t.

      Sigh.

  3. david searles says:

    Anyone have an idea on future sales life on the 6500 series and is
    vPC going to be avail­able or only VSS on the 6500s
    Is there going to be a small 7K box avail­able ?
    My work (col­lege) uses 6509s and 6513s for port dens­ity aggreg­a­tion  — we will con­tinue to need to provide 500 ports @ 1001000 not 500 ports @ 10 Gb.
    A 10 story dorm/​residence hall very likely has 2000 ports and has 4 6513s. Currently we use HSRP with mul­tiple vlans and run 1 Gb & 10 Gb fibre from the 6513s to data centre.

    DS2

Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!