We’re looking for a new compute cluster to replace our aging 4 UCS blades, and I’m hoping some knowledgeable folks can weigh in and offer some recommendations on which to choose.

So far we’re thinking either new UCS blades for our existing chassis, Dell VxRail, or HPE Synergy, but we’re open to considering other options. Not using VSAN, so storage isn’t a consideration.

@Dell_Technologies @Cisco @HPE

4 Spice ups

Compute requirements? Intensive VMs? Compute only, no storage?

Storage IO and size requirements?

Network connectivity requirements?

It’s a little hard to recommend anything while knowing next to nothing.

Run a LiveOptics scan and determine what your environment is doing now. Use that to identify bottlenecks, or areas for improvement. Use that to determine what to purchase.

2 Spice ups

Sorry if I wasn’t clear. I’m not looking for recommendations spec-wise, just for opinions on the platforms themselves. Stuff like hw/sw reliability, ease of use vs. feature set, business practices, etc. Basically, if you really like or dislike a specific platform, I’d like to know why.

If it affects considerations, we currently have 4 UCS blades running on a FC SAN. Each node has dual Xeons and 256GB memory, no local storage except for boot SD cards.

Cisco loves blades. Historically. Up to now before VMware released vSAN Max they didn’t have a product to play well with this Cisco’s idea.

(1) Yes, you can add more blades as a compute and add capacity focused vSAN Max as a storage replacing aging FC SAN.

(2) You can ditch both blades AND FC SAN and go hyper converged all around. We prefer Dell R7xx line of the servers, never had any love to Cisco.

2 Spice ups

Thanks for the reply, but as I said above, we’re not considering storage. We’ve got a brand new all-flash SAN hooked up to our ESXi cluster and a few other systems.

Any particular reason you prefer Dell over Cisco?

No love for blades here. Unless you’re at the scale where multiple blade chassis makes sense, then it’s probably not a good investment.

I haven’t had any unsolvable server issues with Dell or HPE in the last 13+ years.

I guess my last Dell server purchase had more early SSD deaths than I’m used to, but they were replaced without any data loss. Anecdotal evidence, single data point, and thus meaningless to your purchasing decision.

1 Spice up

Like I told Cisco has a focus on blades, read - scalability, and we mostly do 2 and 3 node hyper converged clusters with lots of RAM and CPU cores packed into lowest possible footprint. Just very different use cases. Plus, Dell has nice OEM program and Cisco don’t.

2 Spice ups