Dell’s PowerEdge VRTX Gets An Upgrade

The past few months have seen some long awaited new additions to Dell’s PowerEdge VRTX system.  In an effort to prevent this from being an “all Dell” blog, I’ve held off writing about these new additions until today.  In a quick summary, the new additions include: 2nd PERC, added support for VMware ESXi 5.5, added support for the M820 blade server and the ability to repurpose blade servers from the M1000e.  Let’s dig into each of these.

2nd Shared PowerEdge RAID Controller (SPERC)

pevrtx_les_00090_openWhen the PowerEdge VRTX was announced last June, many datacenter customers wondered why is there only one Shared PERC?  Sure, a single RAID adapter works on a rack server, but that’s a single server.  With VRTX you are able to support up to 4 servers, so a single PERC wasn’t satisfactory as it became a “single point of failure.”  The reality of the situation is there was always a plan for a 2nd PERC (as exhibited by early marketing photos showing dual PERCs) however Dell encountered unexpected engineering delays that pushed out the release of the 2nd PERC until last month. 

The 2nd PERC is set up in an Active/Passive design so that if the primary PERC fails, the secondary PERC will assume the load without any data loss.  Ideally this resolves the concerns about a single PERC being a single point of failure.  One thing to be aware of that seems to be catching some people by surprise.  In an Active/Passive design the PERC’s cache is shifted from Write-back to Write-through which may cause a slight decrease in performance (depending on the workload.) The 2nd PERC is available to order at point of sale or as a customer-installable post sale kit.  However, be advised that it requires adding an additional SAS riser board, and SAS cables as well as the 2nd PERC so consider getting an installation service so help get it in place.  Here is a list of post-sale part #’s:

Dell Part # Description

386-BBBM

Dual SPERC8 Upgrade for VRTX 3.5 Chassis, Cus Kit, RECOMMEND SVC INSTALL SKUs 996-3219 and 984-1037

386-BBBN

Dual SPERC8 Upgrade for VRTX 2.5 Chassis, Cus Kit, RECOMMEND SVC INSTALL SKUs 996-3219 and 984-1037

Demartek recently published a technical white paper on their findings with the 2nd PERC.  I encourage you to take a few minutes and check it out: http://www.demartek.com/Reports_Free/Demartek_Dell_VRTX_SPERC8_Failover_Evaluation_2014-05.pdf (1.5MB, PDF)

Support for VMware ESXi 5.5 (Finally)

One of the initial use cases for the PowerEdge VRTX focused on extending virtualization to remote environments.  At launch, Dell supported VMware ESXi 5.1 however shortly after VMware released 5.5.  Unfortunately, while Dell added support for ESXi 5.5 on the PowerEdge server family, VRTX was not included.  (From what I understand, the delay was associated with getting the Shared PERC driver to work with ESXi 5.5.)  Although long awaited, support for ESXi 5.5 is now available for use with both a single or dual PERC design.  Here’s the link to the Dell / VMware ESXi 5.5 ISO which includes the driver for the PERC.  Getting VMware to recognize the PERC requires no additional steps if you use the ISO – it shows up as “Shared PERC 8 Mini” on the host’s Storage Adapters tab.  Once the imageShared PERC is seen, then you add storage (Configuration tab > Storage > Add Storage) and you are all set.  If you haven’t seen how ESXi sees the Shared PERC here’s a screenshot.  Dell published a nice whitepaper titled “Configuring Dell PowerEdge VRTX shared storage for VMware vSphere Environment” designed to assist in helping you install ESXi on VRTX. 

Support for PowerEdge M820

imageThe Dell PowerEdge M820 is a full-height, 4 CPU server based on the Intel E5-4600 v2 CPU supporting up to 12 cores per CPU.  Your first thought may be, who needs a 4 CPU system in a remote environment?  The answer is – not many people,  however if you consider using VRTX within the datacenter for Test/Dev or for specific projects, the M820 may be more appealing.  The PowerEdge M820 specs include:

  • up to 4 x 12 core Intel E5-4600 v2 CPUs
  • 48 DIMM slots
  • 4 x Hot-Swap HDD bays (available only to the server node)
  • Up to 4 x 1/10GbE NIC ports (LOM)
  • Connectivity to up to 4 x PCIe slots on the VRTX chassis

As you may notice in the image above, the PowerEdge M820 is a full-height form factor so when used it will reduce the server expansion to another M820 blade or 2 x half-height compute nodes.  [The VRTX image shown is the rack version.  It is available as a tower model as well.]

Support for M1000e Chassis Blade Servers

If you didn’t know, the servers used with the Dell PowerEdge VRTX are the exact same blade servers that Dell uses in the PowerEdge M1000e chassis.  The difference between a blade server used in the M1000e and one used in the VRTX is the firmware and the addition of PCIe mezzanine cards in the mezzanine card slots (allows the blade server to communicate with VRTX’s PCIe slots.)   So, could you take a blade server from the M1000e and put it in the VRTX?  Up until recently, you “could” but a) it wouldn’t be supported and b) you couldn’t buy the PCIe mezzanine adapter cards.  Finally that has changed.  After numerous requests from customers and from Dell sales teams, Dell now offers customers the ability to purchase the PCIe mezzanine cards that will allow a M520, M620 or M820 blade server taken from the M1000e to work within the VRTX chassis.  This process is not very complex and doesn’t require (or recommend) any installation service.

Here is a list of post-sale part numbers:

Platform

Dell Part #

Description

M520, M620

330-BBBM

VRTX PCIe Pass-Through Mezzanine Adapter, Quantity 2, Customer Install

M820

330-BBBV

VRTX PCIe Pass-Through Mezzanine Adapter, Quantity 4, Customer Install

Before you ask – I’ve heard there are no plans to add support for the PowerEdge M420 or the EqualLogic PM-4110 storage blade.  I don’t know why but I don’t think many people are losing sleep over that decision.

As you can see, Dell is continuing to invest into the PowerEdge VRTX chassis with more coming later this year, including the addition of next generation blade servers.  If you have any questions, or would like to know more about the PowerEdge VRTX, please leave a comment below or reach out to me directly at kevin AT bladesmadesimple.com

 

Kevin Houston is the founder and Editor-in-Chief of BladesMadeSimple.com.  He has over 17 years of experience in the x86 server marketplace.  Since 1997 Kevin has worked at several resellers in the Atlanta area, and has a vast array of competitive x86 server knowledge and certifications as well as an in-depth understanding of VMware and Citrix virtualization.  Kevin works for Dell as a Server Sales Engineer covering the Global Enterprise market.

Disclaimer: The views presented in this blog are personal views and may or may not reflect any of the contributors’ employer’s positions. Furthermore, the content is not reviewed, approved or published by any employer.

3 thoughts on “Dell’s PowerEdge VRTX Gets An Upgrade

  1. Bede Carroll

    Any hope that the SPERC will support some form of JBOD? Think this would be a great platform for VSAN but the idea of creating a RAID0 for each disk seems unnatural to be honest.

  2. Jose Hernandez

    The performance hit from not having the write-back enabled in the dual SPerc8 is not trivial at all, unless your workload is heavily tilted towards reads, which seems to be the case with how Demartek conducted their testing. In a nutshell, write performance is simply abysmal as Erik points out in his thread. As it is today, if you intend to run a vsphere cluster on the VRTX, the dual SPERC8 setup will not perform.

Comments are closed.