If you haven’t heard yet, VMware announced VMware vCloud Director (vCD) last week and it is gaining a lot of buzz around the industry.  The simplest way to think about vCD is in the terms of layers. It will sit on top of vCenter and extract all of the resources that vCenter would manage and pools them into large aggregates that can be carved up based on resource needs. A good real world example of what this could be used for would be multitenant hosting, or large use of resource pooling internally (think private cloud).  

Some basic things to keep in mind:

  • This works in conjunction with vCenter and vSphere.
  • You will want to use this with VMware vShield, vCenter Chargeback and vCenter Orchestrator (vCO) for full benefit.
  • There will be a self-service portal that will allow for micro level deployment of applications, virtual machines and other IT resources. Ultimately creating a vApp, more details below.
  • It can be clustered with the addition of multiple vCD servers. This will give an incredible amount of scalability.
  • Currently there are three types of resources that can be allocated by a tenant 
  1.                 Computing: Resource Pools
  2.                 Networking: vSwitches etc..
  3.                 Storage: VMFS and NFS shares

Value and Benefit:

High-level, this would be used to reduce the amount of time it would take to deploy VM’s, vApps, and templates. It would reduce the amount of IT tickets and most of all time for all the different groups to come to a consensus on how and what to carve out. Once the established policies are in place each department could log on to their portal for the prompts and have a working VM with the needed software in minutes vs. hours or days in some cases. They call this Self-Service for the cloud.

vCloud also fits in well with some of the primary hardware vendors. There is a documented relationship with Netapp, EMC, Cisco and HP. All have different flavors and benefits; however all have the same goal of allocating and pooling hardware resources on top of the virtual layers. Picking the right partnership could be challenging based on all the options and choices. But you should have comfort in knowing that some of your existing infrastructure will work with this technology, HP,EMC etc…

This is really an exciting announcement, not only because it adds another layer of resource pooling but because it is starting to fit right in line to the converged infrastructure conversation. Cisco was to first to market with their Unified Computing Servers, then HP with their Matrix options and now VMware. There is going to be a ton more of innovation with this concept and it will affect all areas of your IT infrastructure; ____as-a-service. More details to follow.

Using Older Hardware with VMware ESX; Make Sure It Is Compatible!

Part of planning a VMware deployment is deciding the hardware that it will go on. A common solution to reduce costs is to use existing infrastructure of older machines and just place VM’s on this platform. While this will work there needs to be some considerations when deploying on existing hardware.

The very first thing to always do is check the Hardware Compatibility List (HCL). This is the supported list from VMware with “guaranteed” approved hardware.

Also, 64Bit seems to be a reoccurring topic. There are strict limitations what will and won’t work and VMware has built in safeguard to limit installation on unsupported chip sets.

Another huge consideration is to evaluate your virtual machine roadmap. For example, your older hardware may be able to support ESX 3.5. If you were deploy based on that, then wanted to migrate to 4.0 or 4.1 you would be stuck buying new hardware and dealing with a migration. Which isn’t bad except for the fact you need to make sure your CPU will be compatible with VMotion.

Also take into account there is a limit to the amount of compatible machines. Below is a quick summary of the compatible HP systems for ESXi 4.1. As you can see the list isn’t too long. Granted this is just one vendor, it does show that there needs to be careful thought when planning out your hardware.

Granted, I do recommend using older hardware when possible. A good example is test/dev environments or a lab type situation. That being said any thing to do with production machines I would think twice before re purposing older machines.

HP Looks to Outbid 3Par In Virtualization Strategy

Happy Monday everyone! Pretty big news today, HP decided to but in a $1.6 billion dollar bid in for 3Par, a virtualized storage vendor. This is pretty big news in the fact that Dell had offered $1.3 billion in what looked like a slam-dunk deal. Of course, the board of 3Par has to approve the offer; I find it hard to believe they won’t pick the larger price of the two. However nothing is final as of yet.

This by itself isn’t much of a story (at least from virtualization standpoint), although HP to acquire such a big company during CEO transition is an interesting side note. But what this does do is showcase the building up of the competitive unified cloud space (private and public). First you had Cisco break he mold with their UCS story. Then HP countered with their HP Matrix. Recently EMC announced the vblock (VMware, Cisco, and EMC storage) and Netapp announced a partnership with (Cisco, VMware, and Netapp Storage). In short, Dell was basically left out of the loop.

It wasn’t a surprise that Dell made the offer, they need something to add to their Equallogic line, and their partnership with EMC (reselling CLARiiON boxes) isn’t the strongest at this point. If Dell doesn’t make a move on something fast, I don’t think they will have a unified solution story, which I would think, would put them at a huge disadvantage in the near future. 3Par is a good target, and I doubt Dell will give up without a fight.

Long term there is a lot more questions than answers. I find all the recent acquisitions to be a bit unnerving, as do many of my clients. It is hard to make a large purchase knowing that your potential vendor is an acquisition target.  As well, you don’t want to miss out on the new hot thing, or worse invest in something that is obsolete in 12-18 months. Until the dust settles there could be a lot of projects on hold until people trust that there solutions will be around for a while.