Cisco's servers are stuck in limbo, look likely to stay there

UCS has a high hardware failure rate and lethargic support, even if you are a large enterprise.

Expensive & not sysadmin friendly

We were brainwashed to buy a couple of Cisco servers a while ago. They're expensive for what they are, include zero support out of the box and from a sysadmins view, they're not the greatest. Remote management requires Flash *AND* Java, and they just refused to boot with the Intel NICs we put in them. (Our Dell servers work fine with said same Intel NICs)

Cisco used to claim having special sauce to enable bigger memory setups, but nowadays, they're just another tin vendor. Against the likes of Dell, I see no reason to buy Cisco servers.

I am Cisco's lack of suprise

Cisco is a network company. Ok, so it also dabbles in clothing and book selling. But there seems to be a problem of getting too big and forgetting your roots. So networking, not trying to dabble in commodity server shifting. Especially when Cisco's competitors have been merrily chipping away at their core networking business.

Quit? yes, yes they should

I have experience of Cisco, HP and Dell servers.

Cisco's systems management model is archaic.

So why pay more for slower hardware, non-scalable architectures with a management and maintenance headache?

Seriously?

Are there any Enterprise Admins out there that are happy with a Dell or HPE solution? I haven't looked at either in years. I do not work for Cisco and I am a customer. I would love to go multivendor in my datacenter, but the random hardware failures I have had in the past from HPE, and the lack of tools to push firmware to vSphere hosts, I cant go back to that. We have too many hosts. What am I missing here? We are 100% blades on the Cisco side. I can't imagine the time it takes to get new pizza boxes racked and configuring the ports on the network.

Rack Servers are a Commodity

As such, vendors like Dell and Supermicro will dominate.

Blades are expensive

Blade infrastructures cost more and become obsolete. I worked on the original Compaq / HP C class design blade systems and it was intended to be UCS killer with less networking features. 1U and 2U racks systems are still the most popular.

Too much reorganisations... to slow to respond to market shift

Having worked for Cisco, I can tell that the introduction of UCS was revolutionary technology. However it came a couple of years too late and had short lived success. Even though I was built for virtualisation, if the market is driving hard on hardware dependence, you should not bet on building such a clear lock-in system.

The tech behind UCS is still very impressive, but the price is way to high and senior and executive management way to stubborn to see this was a problem.

When I joined it was a company of network engineers and a network sales. That changed and more people with datacenter, software and server knowledge came in. However, I spent a lot of years working for Cisco, but in my time there were a lot of reorganisations...

Sending off thousands of people including a lot of people with the right knowledge to please the shareholders is not the right way to change the company for the best. Not that the 80% network-based sales compensation plans helped.

And then one day I found out that my full team was going to be decommissioned. We held a lot of the datacenter and server knowledge... most people, including me took the severance package.

Short term result was a lot of deals were lost. Long term was that all the long term, read too expensive to lay off, mostly network trained personnel stayed... IMHO this is what will kill Cisco. It is a real shame.