XS23 Cloud Server

There has been some recent press around some of the equipment we’ve developed in our cloud computing group. The core of our business is essentially a consulting and design service and developing new products for customers is a big part of the fun. Because these aren’t mainstream PowerEdge systems, we don’t get the chance to show them off as much as we’d like. Our group has been talking for some time about “optimized designs” for cloud and hyperscale computing without showing what that can really mean, so it’s time to unveil something that’s come out of the lab.  Pictured here is one of our favorites: the XS23.

image

XS23 front – twelve 3.5” SAS or SATA drives; 3 per server

This product was designed for a customer that needed maximum compute density, a healthy amount of local disk and, of course, lowest power draw possible. Our architecture team threw all that in the blender and out came a 2U standard rack mount chassis that houses four dual-socket servers and twelve 3.5” hot plug drives.

image

XS23 exploded view: two dual-socket servers mounted in chassis bottom; two in a mezzanine above. Industry standard rack-mount chassis.

Density of this type is certainly not unheard of (half depth or twin 1U’s), but by going to a 2U chassis we were able to fit it with larger, more efficient fans and stack 3 rows of full 3.5” drives across the front. So, even with a 25% higher density than general purpose blades, it provides three local spindles of 3.5” SAS/SATA disk to each server. Of course there are tradeoffs. This was expressly designed for an environment with high node failure tolerance - a cloud application. By designing out a lot of the capabilities that weren’t required (like redundant power) we were able to deliver the performance and power profile required. Efficiencies are gained by shared resources - as seen in a lot of general purpose designs available today. We think the key to designing the perfect cloud server is knowing where to stop and also what not to build in. This is a function of each customer’s unique design goals. Applications truly capable of foregoing high availability in hardware are somewhat rare, but customers in this space have it – as well as a laser focus on their business levers. So in this case we took the problem statement and made the tradeoffs to yield highest efficiency and density within the performance parameters of the application.

It’s important for me to emphasize that the XS23 is not generally available. This system is qualified and supported for only a handful of specific customer applications and locations; it’s not completely productized to bear a PowerEdge badge. I hope you’ll watch this space for more unique designs and the discussion on cloud taxonomy and architecture that Jimmy's leading.

Comments  Comment RSS Feed

Michele said:

So how many servers would a company have to get in order for Dell to this sort of custom build?

 

Michele 

Todd Brannon, Dell Data Center Solutions said:

Hi Michele,

Fair question.  There are definitely some economic variables in our approach.  In terms of unit quantities, if we are leveraging an existing reference design it can make sense for as few as 500-1000 systems.  For a blank-slate co-development effort the scale is typically 1500+.   Much of this ties to the logistical elements of manufacturing, deployment and support. 

Aaron said:

Wow, this is awesome, and looks like it deserves more attention.  Pretty good article about it as well.  Great work guys!

Jason Wallis said:

Can you quantify the power savings (in general) of the XS23 over deploying 4 standard dual socket servers with redundant power supplies but similar local storage and computing power?

Todd Brannon, Dell Data Center Solutions said:

These weigh in with savings similar to what you see with blade systems.

John said:

Why have the local drives at all? 3 drives per machine in what I expect must be a Raid5 arrangement means less than 2tb local. 

What are the HBA connectivity options? Why wouldn't a cloud infrastructure rely on SAN or iSCSI based storage?

Have you ever done a bare bones build out of these tuned to run ESX or something along those lines? The R900 is a beast and I can't wait to take delivery of three of them in the next few weeks but this clearly seems like an order of magnitudes beyond that?

Would it be possible to take some of the redundancy out of the R900s and make them more energy efficient if they deployed to an ESX HA enabled site? 

 

 

Todd Brannon, Dell Data Center Solutions said:

John - good questions all.  The local storage was a requirement from the customer.  We do have options for FC HBA's on this system but interestingly enough most of what we see in the cloud computing group is ethernet only.  Some of our customers with more traditional HPCC type environments use SAN out to every node; in the big search/web2.0 clouds it's much less common.

The XS23 only has 8 DIMM slots-  it's not designed for a tall virtual machine stack to the capability of a system like R900.  This sytem was point-developed for one of our large cloud computing customers with a home-grown application.  As customers deploy systems into environments that don't have as much need for HA we'll likely see designs like this become more mainstream.

Leave a Comment

Compose
Preview
(required ) 
(required , not published) 
(optional )
(required ) 

Note: Conversation is encouraged and expected. However, moderation of comments is necessary to prevent spam, personal attacks, profanity, mentions of legal action or off-topic commentary. We will not publish comments that advertise third-party shopping sites or ones that violate our terms of service.

Comments related to specific product support or customer service issues will be addressed separately rather than posted here. Please use the links in Contact Us for product and customer service assistance.