My thoughts on the major announcements from VMworld 2012

I believe the theme for 2013 will be, “The year the hypervisor becomes a commodity.” Microsoft, Citrix and KVM are gaining enough parity with ESXi that they will be good enough solutions. I guess a better theme will be “2013 the year of the good enough hypervisor.” I was expecting major announcements on the management front from VMware along with continued strides on hypervisor innovation. VMware did not disappoint in either category.

First on their core business the hypervisor, I thought the announcement on the death of vRAM entitlement was a tad uncomfortable. vRam was a mistake that should have been corrected right after it was made. It was no surprise that the first announcement was the death of vRam entitlements. I don’t consider this innovative but rather a prerequisite to any continuation of the vSphere product line when looking at hypervisor competition. The fact that VMware’s newly announced CEO Pat Gelsinger had to ask for bigger applause indicates I’m not the only one that felt this way. There was also continued innovation on storage. VMware announced the anticipated virtual SAN which allows SMB to pool the local storage of ESXi hosts to create a virtual SAN. Also, VMware announced enhanced vMotion which allows the migration of virtual machines sans a SAN (pun intended). VMware said that enhanced vMotion would also support DRS which is a cool trick that I’m anxious to see.

These changes all lead to the continued drum beat of the virtualized data center which was branded by VMware as the Software Defined Data Center (SDDC). VMware wants to own the data center from server to storage to network. VMware has preached the virtualized data center since vSphere 4.1. Now that they have acquired both DynamicOps and Nicira, I believe VMware can actually be taken seriously. These two acquisitions combined with the announcement that they will be a gold member of the OpenStack project helps lend credence to their claim of wanting to run the data center. I have questions as to how VMware’s strategy will mesh with OpenStack as vCloud, vFabric and, vFoundry are all direct competition to Openstack.

VMware showed the strength of their vision with application aware vCloud 5.1. I was extremely impressed with the vCloud Hadoop demo. As part of the demo VMware showed how not just the infrastructure is elastic in vCloud but the elasticity stretches to the application stack. They were able to provision Hadoop application and database nodes that resized the Hadoop stack. I don’t know how much of this was real versus story boards but the claim is very impressive.

I’m not too excited about VDI in general. VMware had announcements around its Horizon solution but I have to be honest the VDI is a niche solution. I’ve shared my thoughts in the past. For those needing VDI they will buy VDI. For the rest of us trying to figure out BYOD we’ve decided VDI isn’t the solution. It may be a stop gap solution until something better comes along for mobile devices. From a traditional since I believe VMware is well positioned with its growing management suite and the fact the Microsoft is shipping Hyper-V as part of Windows 8 to do some very interesting things on the full client side with portable offline mode VDI.

I’m also pretty excited about VXLAN or should I say vXLAN. It’s the most enterprise relevant part of the whole SDN movement I’ve seen so far. The demo of extending your data center virtually to public clouds using the same address space was network geek nip. This is another demo that I have to see out in the wild. The thing about software demos versus hardware demos is that you can make anything look smooth with software in a controlled setting. That is unless you are Bill Gates.

I did VMworld from afar again this year. I promise I’ll attend in person next year and hopefully do some live blogging.

Published by Keith Townsend

Now I'm @CTOAdvisor

7 thoughts on “My thoughts on the major announcements from VMworld 2012

  1. VXLAN is not a solution for extending to public clouds by itself. There are too many timeout counters that people forget about for it to work towards a public cloud. As Allwyn Sequeira, VMware CTO, described during the SDN Networking and Security session, it is a local datacenter solution to overcome the limitations of the amount of VLANs required in multi-tenant environments.

    Remember VXLAN is multicast based;
    – Can you imagine requiring native multicast support over the Internet towards the cloud providers? Imagine the required routing tables for those S,G trees!
    – Imagine how many ARP request timeouts you will have trying to do an ARP resolution across the Internet?

    VXLAN is one technology that solve a huge problem for multi-tentant environments but it is not the solution for everything, and for sure, not for “across clouds” extension of the networks.

  2. Thanks William. Your breakdown makes since to me. I have to go back and look at the Keynote. I thought the presenter mentioned VXLAN when he was extending out to a public cloud. It would make since if the private cloud had a private connection to the public cloud provider. However, the public cloud provider supporting VXLAN would be a stretch at this point in the technology’s maturity.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: