blog.virtualtacit.com

I don’t know about you, but I cant speak enough about the fourth generation CX. EMC engineering has really out down themselves by producing a super sexy, ultra-flexible, power efficient juggernaut.  Code named, Fleet, the CX4 brings some major design changes as noted here. Architecturally, UltraFlex is the spits, the driving factor for customer investment protection….

Why? See below–

By separating the processing portion of the SP’s from the I/O connectivity portion, throw in a pinch of software intelligence along with hot-pluggable, non-disruptive I/O modules and you have the makings of a pay-as-grow highly scalable SAN, second to none. No more special “C” units (CX3-20C), notably in this line all base configs will have FC and iSCSI ports, that’s right they’re all combo units. And with the advent of 8Gb FC and 10Gb iSCSI as well as FCoE, you can be assured that this line will accommodate future technologies without a forklift upgrade.

Also new to this line is Flare Code 28, the first 64-bit OS for the CX. Here are some of the new features specific to Flare that weren’t mentioned previously..

• Improved cache management-The improved cache management feature allows non-mirrored single storage processor write cache operations to continue when the peer storage processor is unavailable. In addition write cache operation is retained during power supply failures, fan failures and non-disruptive upgrades (NDU).
• IPv6 management ports (yeah I know who cares)-You can configure the management ports to run IPv4 or IPv6 using Navisphere Manager.
• 1 TB SATA II drives in first enclosure for CX4-120-The 1 TB SATA II drives are supported for the first five drives in the first enclosure for the CX4-120. The CX4-120 is the only CX4 that supports the 1 TB SATA II drives for the first five drives in the first enclosure; the other platforms require Fibre Channel drives. The 1 TB SATA II drives are supported as general purpose drives on all platforms.
• RecoverPoint-The splitter component of RecoverPoint allows you to use RecoverPoint when no server-resident splitter is available for an attached server.

What? you don’t believe me well here is what others are saying..

Straight from Sakac’s mouth are a number of valid points on “VMware’s  IO centric Strategy” and how the CX4 is a catalyst for this proliferation, http://virtualgeek.typepad.com/virtual_geek/2008/08/our-vmware-cent.html.

Mark Twomey over at Storagezilla has done a great job of covering the in’s and out’s of the new line and what key features present this line as a formidable force to be reckoned with, http://storagezilla.typepad.com/storagezilla/2008/08/cx4-ultraflex.html.

And last but not least, virtualization.info has briefly touched on the introduction of this CLARiiON to market (although mostly phrasing from Sakac’s site) http://www.virtualization.info/2008/08/emc-introduces-massive-vmware-support.html.

The point here is, competitively, EMC has now positioned itself with the CX4 as a trendsetter against the likes of IBM’s DS4000 series, HP’s SW EVA and NetApp’s FAS series. For these arrays to rival they must have I/O modularity, FC and iSCSI natively (Note:NetAPP does offer in FAS series but not natively-WAFL- therefore performance suffers), and above and beyond integration with VMware. So if you are questioning which array is right for you make sure you patter these questions against your array of choice and see where current feature sets add up:

• Do you have VMware in your environment if so is my array “virtualization friendly”?
• What is your current DR solution and how does your storage vendor create opportunities for improvement in this arena?
• What specific application level requirements do you require, can your array spell QOS?

The average run of the mill midrange array it isn’t…agreed?