Shifting to Software-Defined
Showing results for 
Search instead for 
Did you mean: 

HPE Synergy and Cisco UCS: Comparing four typical datacenter tasks by the numbers



Tom Cloyd.jpgBy guest blogger Tom Cloyd, Senior Marketing Manage, HPE 

When you are setting out to evaluate different datacenter solutions, it helps to focus on the major items that impact operations.  Simply put, addressing recurring activities is the most impactful way to drive significant savings in time and efficiency. By reducing the number of steps in a reoccurring task, your time savings is realized each time you do that task. Of course, the more steps you have in routine tasks, the greater the opportunity for error.  Therefore, reducing the number of steps can also remove significant elements of risk from your datacenter’s day-to-day operations.

There are four server (compute node) recurring activities that come to mind:

  1. Deploying new servers
  2. Updating firmware and OS patches
  3. Taking advantage of VMware vSAN
  4. Setting up Boot from SAN

HPE asked Principled Technologies to do comparison testing for HPE Synergy and Cisco UCS on these four activities, looking at time to complete them, how many steps it took to do the work, and how many different tools (consoles) were needed.  You may find the results surprising….. “HPE Synergy composable infrastructure requires fewer steps and less time than Cisco UCS” 1.  table 1.png

HPE Synergy is less complex?  That’s exactly what the table above shows – and does so dramatically:  fewer steps, fewer tools, and faster. Welcome to the trifecta of server management.  Not exactly what you have been hearing? Let’s look a bit deeper.

If you put together the results from all four use cases, the HPE Synergy solution saves a total of 71.5 minutes and 69 steps; and as a bonus, Synergy uses 4 fewer tools than the Cisco UCS solution.  The savings columns in the table above tell the story. Synergy really does take much less time and many fewer steps than Cisco UCS.

Sure, in some instances the time required may not be that different for doing just a single server – such as enabling Boot from SAN (only 4.5 minutes difference). But the difference in the number of steps required (31 more steps for Cisco UCS along with 31 more times opportunities for errors) is a bit scary.  Remember, this number is to do just a single server set up.

While the time savings for a single server, or a single use case is interesting, what is really compelling is how this plays out at scale.  Depending on your point of view, it could be disconcerting.

In today’s world, security is paramount. You don’t have to look too hard to see that fact play out in the news in real time.  Multiple elements play a role in keeping your datacenter and your information safe. Some of the most basic include ensuring that every OS is as up to date as possible (in order to plug known vulnerabilities) and making certain that the latest firmware is in place.  The table above shows the single instance comparison between HPE Synergy and Cisco UCS, but that data is only part of the picture.  The full story, as mentioned earlier, is what happens at scale. 

table 2.pngTake a look at the following two graphs. The first is for 1 compute node (server) up to 100 nodes.  What you see here is the time it takes to update the firmware and apply OS patches for ESXi hosts.  Remember, this shows the time is takes to do this step just one time across these server counts.  Notice that you really don’t need to have a lot of servers for HPE Synergy to make a difference for your business. 

Most datacenters do a single update a year in the best of circumstances.  Historically, as you can see from the Cisco UCS numbers, it takes too long and with the number of repetitive steps involved, the process is subject to errors.  The potential for fatigue errors should also be considered when staff is performing large numbers of repetitive steps.  Looking at the 100 node numbers, Cisco UCS takes almost 45 hours to do this work.  That’s over a full week of an IT admins time spent on updates, not to mention the 1,600 steps that need to happen as part of the process.  Contrast these numbers to HPE Synergy only needing 17 hours and 1,200 fewer steps.  And this is just for 100 servers. What if you need to make multiple updates and patches during the year?  The lost time hurts, and the element of risk increases rapidly.

But what if you have a larger datacenter?  Here is how that plays out.  At 2,000 servers, Cisco UCS is approaching 900 hours to do a onetime firmware update and OS patching--with 32,000 steps! 

table 3.pngThese numbers are more than a bit daunting and explain why many data centers don’t like doing updates.  Look at it this way.  There are 2080 work hours in the year.  Now take out 4 weeks for vacation and another week for holidays.  Now you are at 47 work weeks, 1,880 work hours. 

With a Cisco UCS solution your business just spent one half of an IT administrator’s time for an entire year doing a single server update iteration across your install base.  Do it twice, and now you have a year spent updating 2,000 Cisco UCS servers.  How does HPE Synergy compare?  The graph tells the story.

With only 340 hours, HPE Synergy takes over 60% less time to do the job.  This is 540 hours less than Cisco UCS – over 3 months less time.  Plus, it takes 75% fewer steps and as a result is easier, less error-prone, and by definition more efficient.  With a Synergy solution, your IT department effectively recaptures over a quarter’s worth of production for a staff member.

Less time, less risk from errors, and more productivity equal greater levels of efficiency and savings in datacenters.  And don’t forget--there is an increased ability to respond dynamically to the needs of the business and the marketplace that accompanies these increased efficiencies.  What’s not to like?

This same scaling scenario plays out across all the use cases that Principled Technologies tested.  An important note on this research paper is that with each tested scenario, for either time or steps and sometimes both, there is always a compelling savings with HPE Synergy. Table 4.png

Take a look at the in-depth analysis and see all the steps involved and the configurations used for both the HPE Synergy and Cisco UCS comparisons.  One item not touched on here, but is covered in the report, is the number of tools that are needed to perform these four activities.  I won’t keep you in suspense, but it’s not a real surprise either – HPE Synergy uses fewer tools than Cisco UCS, up to 66% fewer in some use cases.  Read the Principled Technologies report here to see all the details.   

Learn more about HPE Synergy online and see what this new cutting edge technology can deliver for your datacenter.  Synergy provides a programmable infrastructure as a service (IaaS) architecture that is managed from a single pane of glass.  This unique design providing a software-defined datacenter (SDDC) lets you consolidate your applications and workloads onto a unique architecture that delivers fluid pools of resources managed by a single tool.  HPE Synergy has set 17 SPECjbb_2015 world records plus 4 new ones by the SY660 Gen10, holds the top VMmark 2.5.x record, and is certified for Oracle VM, SAP HANA, and VMware vSAN – just to name a few.

To learn more about HPE Synergy, download the free e-book: HPE Synergy for Dummies.

And click here for the HPE Enterprise Library.


Follow HPE Composable Infrastructure

Note: all the data points used in this blog were rounded down at the single server or use case level.  What this means is we have understated the HPE Synergy time savings discussed above.

1 Principled Technologies report Deploy and manage servers more efficiently with HPE Synergy; August 2017

2 Modeling was based on the referenced Principled Technologies research and done at 2,000 nodes for scenarios 1, 2, & 4, and at 2,000 implementations for scenario 3.

0 Kudos
About the Author