Cat 4500E Sup 9-E, To Be, or Not To Be

casanavep · ‎10-19-2017

Is it just me or is the release of the Sup 9-E the quietest 4500 leap in the history of the platform? Is it intentionally being kept quiet during its initial limited release (only shipping to select customers right now) in an attempt to keep customers focused on the Cat9K releases; or, are there issue in this limited release that has kept Cisco from doing the floods of YouTube and TechwiseTV videos that they have during all their other launches over the last few years? This part seems like a huge breakthrough for the investment protection at many companies whose IDF (and small campus core) switches need not scale beyond 48Gbps per line card, but still want the SD Access / DNA road map, converged wireless and wired access, and futureproofing that newer programmable ASICs bring. I am just confused on why Cisco would be so quite about a product that has already started its launch process, outside of possible marketing tactics or a launch with issues. Anyone have any actual insight on the absolute lack of marketing and social media hype towards what appears to be a potentially significant impacting part for much of the SMB campus enterprise community?

Joseph W. Doherty · ‎10-19-2017

As a guess, I would suspect Cisco might not what to compete against itself especially when the new Cat9k might move you into buying a whole new chassis and components, not just a new sup.

Leo Laohoo · ‎10-19-2017

The biggest problem with those who purchased the 1st generation of Sup7E bundle is the chassis "version". Early units of the Sup7E chassis are unable to upgrade to Sup8E because of this limitation.

Sup9E removes that restriction. So this means those Sup7E with chassis version of 1.1 can upgrade to Sup9E.

But I do agree. What is the "benefit" for upgrading to Sup9E when the 9400 is meant to replace the entire 4K line.

Reza Sharifi · ‎10-19-2017

Leo,

If I had to upgrade from 8E, I would defiantly go to 9E before touching 9400 series.

I know for a fact that the 4500 series work and they work well but can I say the same about 9400? (:

Reza

casanavep · ‎10-20-2017

Leo,

To me the benefit would be investment protection. Let's be honest, 99% of the worlds campus wiring closets don't need more than 48Gbps per line card; so, if their core and only desires to upgrade beyond Sup7e were based on SD Access, DNA Automation, and the security it brings, hat did going to a 9400 bring to the table that the Sup9-E doesn't appear to be? I am guessing that if you have 100+ of these to replace globally, a huge hit to your IT budget and manpower hours. I worked with companies with 1000's of call center seats globally. This low margin work is a perfect example of where we can benefit from the security and automated predictability without the extended downtime and cost of swapping full chassis and their associated linecards. Another customer base I work with is the US academic, where again even the most digitized schools are not pushing the 48Gbps per line card cap outside of their data centers and research facilities, already have linecards that do UPOE with deep CAM resources, but always need to more tightly secure the security aperture on extremely tight budgets. There are tons of other examples where the benefits of programmable ASIC based Sups with deeper tables for SD Access overlays will open up new world for businesses, but are priced out in human and capital resources. The Sup9-E seemed to be a great path for these organizations. In the end, these organizations end up having to delay upgrades due to finite resources, all too often because excessively sized upgrade paths like the 9400. I am not saying it is not a great switch, but like the Lamborghini, more than many need in their wiring closets for the foreseeable future, especially if the 4500R+E with Sup7-E and UPOE linecards was just pushed to them less than 5 years ago as the way to go. I try to treat their money and human capital as my own when making designs. If I cannot honestly convince myself that a solution is right based on quantifiable metrics vs new toy appeal, I won't try to convince a customer on it.

So, I believe the answer to your question as to "what is the benefit" is getting my customers the features the want and need (even if they don't yet know or understand it) at the best price and supportable path possible. My question is the same but in reverse order, given the realities for the customer groups I mentioned above, "what is the benefit of Cat9K over Sup9-E" beyond longevity? I say beyond longevity because I have seen the new thing turn to old faster than we thought, think CIUS.

Leo Laohoo · ‎10-20-2017

@casanavep wrote:

To me the benefit would be investment protection.

Before I begin, I'm not here for a debate. Reza has a point. Joe has good inputs too. Here goes ...

This first line caught me. I used to work in an environment where the mindset was very "risk averse". We were happy to buy Cisco routers and switches even though EoS has been announced. More so, we were happy to run IOS codes that were way, way out ... Because they were deemed "stable".

IT budgets are hard to come by. When it comes to networking, we see the entire IT budget get sucked/squandered away by more glamorous stuffs and leaving upgrading networking kit picking up the crumbs (after the sweepers have gone through).

Right now, I am very lucky to work in an environment where the mindset is the complete opposite. I have implemented a few stuff (routers, switches, wireless AP and controllers) with the models just recently announced or, in some cases, soon-to-be-announced.

We understand software bugs (aka "undocumented feature") and we've hit them. But none of them are show stoppers. We've either raised the issue with TAC and/or we've found workarounds. We regularly (several times a year) upgrade the firmware of our routers, switches and wireless controllers.

Joe and Reza has seen them all. They've been in the "trenches" and the firing line more than I. Their reasons and justifications are more than valid. This is not a one-size-fits-all.

If you've seen a lot of my posts, I would always recommend people to purchase the newer products over the existing models. I, personally, don't see the real benefit of buying something which is (very) close to EoS than getting the model that supersedes the current one.

casanavep · ‎10-21-2017

Thanks Leo Et al. I haven't seen the EoL/EoS announcements for 4500R+E series chassis, just Sup6 and Sup7; so, was under the assumption that Cisco would follow their standard protocol of a 4+ years from announce to EoSupport. Like you, in the 20+ years doing this with Cisco have recommended new when it is an option and won't hinder an institution from making other IT needed adds or changes, or is absolutely needed. If I missed it, and there was an announcement, then I agree. That would be more than enough ammo to stick a fork in the 4500R+E's. The fact that they just started limited release of Sup9-Es makes me think we are still at least 2-years out from such an announcement, which would typically mean at least 4-years of support window beyond that. Again, I could be wrong and have missed something announced. If I am not, and we likely have 6+ years of supportability and this being build on Cisco latest ASICs pretty much guarantees latest code flavors / enhancements, I am still stuck trying to justify, as you said, possibly "squandering" those rare dollars for my own comfort or simple desire for new vs real need. Thanks for keeping my thoughts in check. Like all of us here, I think we just want to give our customers the best possible outcome, while providing fiscally responsible ability to scale and grow their institutions.

Cheers,

P