Home > VMware, VMworld > VMworld Buzz: Day 1

VMworld Buzz: Day 1

October 18th, 2011

VMworld kicked off today in earnest with about 6500 expected attendees. My day kicked off trying the local cuisine. A Danish pastry for breakfast in…Denmark…perfect!

The morning’s major announcement was all about Management with VMware announcing vCenter Operations Management Suite.

I then attended VSP3067 — Mythbusters Goes Virtual – Eric Sloof and Mattias Sundling.


This was a great session with a few very simple yet very important things that every VM admin needs to know. Busting the myths was all about looking at things like RDMs vs. VMFS, Changed Block Tracking, Resource Pools, and PVSCSI vs. LSI Logic. Eric and Mattias went through each one and explained the things that you may have taken for granted but are no longer true. Even if you don’t have a chance to hear the repeat, look up the content and make sure you are following these great tips.

I felt very stupid as I was sitting next to probably the great collection of PowerCLI minds in the universe!


Here’s a VMworld TV link about the session.


VMworld TV has kicked off as well with Dr. Evil 🙂 missing but his stand-in is Mr. Security, Matt Northam.

Also seen were Monster VM and Chris Colotti spotted him doing a lab too!


I then attended CIM1264 — Private VMware vCloud Architecture Technical Deep Dive – Chris Colotti, David Hill. These are two of he uber vCD gurus so was great to hear them talk.

The presentation started off with talking about how chargeback in a cloud environment brings business agility. vCD architecture was then explained in terms of providers and consumers. Chris and David then formalised the best practice for having a separate management cluster. This is so all your infrastructure stuff is separate from your actual cloud implementation so your VM admins don’t mess anything about within your Cloud as this is really a separate infrastructure and you certainly don’t want a vCenter admin making any resource changes in your vCD. Another best practice was to always use whole clusters for your provider vDCs rather than resource pools. This again reinforces that vCD is really for a very large environments that can justify the management and cost of separate clusters for management and each provider resource vDC. Allocation models were then explained and then what vCD is all about and what makes it so complex which is Networking. In fact Chris said that one of the reasons vCD is so complex is because there are so many networking options! So, network pools was the deep dive part of the session using the standard VMware pictures showing all the network options which are actually good diagrams to get to know. They also mentioned an interesting network concept that when moving to a cloud with vCD you should avoid the temptation to take all your existing VLANs into your cloud but rather only keep those as external networks and have true separation between your cloud network and your external network where your storage, AD, DNS, databases etc. are located. The thing with vCD is it takes a serious amount of time to get to understand the network possibilities so hopefully people in the session had at least a basic understanding of the networking aspect of vCD otherwise they would have left dazed and confused!

Interestingly the presenters then went on to say that for a private cloud implementation vCD is not really for full production use due to the lack of integration with backup and DR products and procedures. They suggest vCD is mainly for pre-production, development, staging, labs, demos and testing. This is interesting as companies wanting to move to private cloud need to understand that when using vCD they can be moving their non-production workflows to it but not necessarily the rest if they can’t back it up or make it available for DR. They went through a use case for vCD and how they used the networking in a clever way for a company that needed multiple DEV environments spun up quickly as vApps but use the same IP address which can be done in vCD.

Next up I attended VSP1883 — VMware vSphere PowerCLI Best Practices – Luc Dekens, Alan Renouf

This was a look at some of the more advanced things you can do with PowerCLI and how to tap into the deep universe of the Get-View command which can grab any possible bit of information out of your infrastructure and went through some of the options for finding out properties using the SDK. Red Hat have been giving away actual red hats and the presenters did have to remind everyone that PowerCLI, being based on powershell was actually a windows management tool!

The presenters then explained that using the Get-View commands can significantly speed up your scripts. Although more complicated to understand in a large environment it may save a lot of time. Alan then talked about scheduling scripts and mentioned a powershell command you can use to create the Task Schedule which sounds great. In the last part of the presentation, Alan and Luc then had 41 different PowerCLI examples which the audience could choose from. We went through about 8 of them and you could see the code for the examples. A great session to review afterwards as all 41 code samples will be available.

Next up was the General Session with Stephen Herrod, VMware CTO.

Some more VMworld TV.

VMware’s End User Computing can change your life, just watch!

A quick look at the new vFabric Application Management Suite

Next up this evening is a vExpert meeting so I’m looking forward to meeting some more people. Then onto some parties…VMworld is busy!

Categories: VMware, VMworld Tags:
Comments are closed.