Home > Blogs > VMTN Blog > Monthly Archives: May 2010

Monthly Archives: May 2010

Top 5 Planet V12n blog posts week 21

You all know the drill by now:

  • Simon Long – Using vMA As Your ESXi Syslog Server
    When vMA collects the logs from your ESXi Host, sometimes the logs have the ESXi Host timestamp and sometimes they will have the vMA Localtime timestamp. I'm not exactly sure why this happens, but it does. (You may or may not know that ESXi uses UTC as its timezone when it timestamps the logs. You can read more about that here. VMware have told me this cannot be changed.)
  • Didier Pironet – The host does not have sufficient memory resources to satisfy the reservation
    I removed the host from the cluster, made a direct connection with vCenter client and booted up the VM, but still same error. I could say that something was wrong with the host only, and not with the VMware cluster’s resource allocation config.
  • Hal Rottenberg – Getting around PowerCLI roadblocks: Tips from the field
    In order to move past this barrier, you have to use the Get-View cmdlet. Get-View is very useful — it's like having a release valve that lets you bust out some really cool tricks. I'm not going to go into great detail about the cmdlet here, but what you need to know is that the PowerCLI cmdlets present an interface to the virtualization administrator which is customized to be consistent with how PowerShell works. This is great for administrators when it works, but when you hit against limits of the included cmdlets, you have to work with the underlying API objects. In other words, you need to understand a totally different interface — the same one that programmers see when they develop software for vSphere.
  • Duncan Epping – Swapping
    As always the common theme of the discussion was “swapping bad”. Although I don’t necessarily disagree. I do want to note that it is important to figure out if the system is actually actively swapping or not. In many cases “bad performance” is blamed on swapping. However this is not always the case. As described in my section on “ESXTOP“ there are multiple metrics on “swap” itself. Only a few of those relate to performance degradation due to swapping. I’ve listed the important metrics below.
  • Frank Denneman – Re: Swapping
    When a virtual machine guest OS starts, there will be a period of time before the VMware tools is loaded and the vmmemctl (balloon driver) is operational. During this timeslot the operating system can access a large portion of its configured memory. Windows systems are notorious for this as they tend to touch every page until it reaches the end of their configured memory. Unfortunately page sharing due to Transparent Page Sharing (TPS) is also at a minimum. Redundant memory pages are not collapsed immediately when a virtual machine is started. TPS is a VMkernel background process and uses a cycle of 60 minutes (Mem.ShareScanTime) to scan a virtual machine for page sharing opportunities.

Top 5 Planet V12n blog posts week 20

It's been a while since the last Top 5 has been released. My apologies, but I attended the VMware Technical Summit, was on a holiday for a week in Italy and two close relatives passed away. Anyway it is Sunday again and tomorrow a fresh new weeks start… no apologies any more. We will continue with the Top 5 as of this week again. It took me a while to go through the immense list of articles, but I guess the following 5 stood out. 

  • Bob Plankers – What are P-states and how do I use them in vSphere? 
    VMware vSphere 4 added the ability to take advantage of Intel SpeedStep and AMD PowerNow! CPU power management features. These features are commonly known as “Dynamic Voltage and Frequency Scaling” or DVFS, and let an OS cooperate with the CPU to reduce power consumption by reducing the frequency of the CPU and the voltage at which it is operating. It reduces these things in preset tiers, and these tiers are known as P-states. On Intel CPUs they are trademarked as “SpeedStep” and on AMD they are either “Cool’n'Quiet” or “PowerNow!”
  • Bouke Groenescheij – Shares: Low-Normal-High
    Well, try to stay away from multi core VMs. If you are running multicore VMs, don't overcommit (at least not too much) and put them on a dedicated cluster. Why not overcommitting? Well, duh, you gave it multiple CPUs, so apparently those VMs need to do some processing which is very important. And since it's that important, don't overcommit. Should make sense. Else give it one vCPU. Also, try not to use the somewhat extreme 'low-normal-high' share setting. I would say: "If a VM is more important, use a custom setting of 1200 shares/vCPU. If a VM is less important, use a custom setting of 800 shares/vCPU". That way the ratio between them isn't that extreme and nature is more balanced.
  • Frank Denneman – Resource Pools Memory Reservations
    If a cluster is under-committed the VM resource entitlement will be the same as its demand, in other words, the VM will be allocated whatever it wants to consume within its configured limit.
    When a cluster is overcommitted, the cluster experiences more resource demand than its current capacity, at this point DRS and the VMkernel will allocate resources based on the resource entitlement of the virtual machine. Resource entitlement is covered later in the article.
  • Scott Sauer – VMware Acquisitions – What’s it all mean?
    There has been a lot of activity here at VMware with acquisitions and partnerships over the past few months. A fellow engineer at VMware summarized a lot of these acquisitions and how they are meaningful to VMware as an organization. I wanted to share this information because I think it provides people with a better understanding of where we are going as a company and the overall strategic vision of VMware (Thanks again Andy!).
  • Steve Kaplan – Calculating the optimal Microsoft SQL licenses for virtualization
    The ability to consolidate multiple SQL instances onto a single host combined with application server license mobility rules generally make either SQL Server Enterprise or SQL Server Datacenter the best choice for a vDC. A 2-CPU VMware vSphere host, for example, running two licenses of SQL Server Enterprise could run up to eight instances of either SQL Server Standard or Enterprise. Additionally, these instances can either be VMotioned to another host in the cluster or even utilize the continuous availability of vSphere Fault Tolerance without requiring additional SQL licensing for the target host.

EMC World day 1 – VPLEX, Joe Tucci & Michael Capellas drop by, and an interesting private cloud TAB

Day one at EMC World kicked off today. The day had what was to me a bit of an odd structure, with a product announcement to the press, a keynote that really didn't go into the announcement, and then two more keynotes and an executive panel in the afternoon. 

Here are some pix from the conference:

The big EMC news today was VPLEX, a new federated storage product that, in its current incarnation, should let you VMotion a single virtual machine or an entire data center across 100km or so. It's new enough – even a new category? — that even the EMCers seem to be struggling to define it succinctly. 

  • Chuck Hollis (EMC): VPLEX: The Birth of a New Storage Platform "Like anything relatively new, it will take a while for people to fully understand the rationale and the strategy behind the product.  It took me a good while before I got a full grasp on the implications of this new technology."
  • Chad Sakac (EMC): Your Virtual Machine Teleporter is ready … Are you? "Why is this important?   Well, one of the key tenants of the “journey to the private cloud” is not only being able to consume things differently (via all sorts of self-provisioning models amongst many things), but also being able to break the barriers of the physical datacenter – being able to do things across geographic boundaries."
  • Stu Miniman (EMC): VPLEX: Redefining the Boundaries of the Data Center. "When new technologies are introduced into the marketplace, people are most comfortable in making comparisons to things that they already know.  In this post, I’ll compare traditional replication solutions with VPLEX Distributed Federation."
  • Storagezilla (EMC): This is VPLEX. "A category creating product which brings Distributed Cache Coherency to workloads in the Private Cloud by providing the ability to make block storage volumes available Always-Active across long distances."
  • Stephen Foskett (Gestalt IT): EMC Shouts VPLEX In A Crowded EMCWorld 
  • Scott Lowe (EMC) has some additional links: EMC VPLEX Launches Today

The bloggers lounge at EMC World was a great place to network and still do work besides. Along with the discussions on storage, virtualization, and private cloud, we were able to watch a succession of luminaries step into SiliconAngle's TheCube for a live video streaming session. All the sessions will eventually get up on YouTube, but for now you can check them out here. For me, the standout session was a surprise visit by EMC CEO Joe Tucci and new Acadia CEO Michael Capellas. This may be the first on-camera intervew from Michael Capellas since he was was announced at heading up Acadia, the EMC-Cisco joint venture. Check out his reasons for signing up to lead this new company focused on private cloud.

Watch live video from The Cube LIVE from EMC World 2010 on Justin.tv

I'd be remiss if I didn't point out that I was also on a live panel discussion on social media. It was a good session, but it really wasn't about virtualization, so if you're interested in hearing us punditize, check out this post at Len Devanna's blog.

Another interesting bit of news via Chuck Hollis was the formation of a Technical Advisory Board at EMC. Nothing very controversial with this group of very very smart people, all of whom I'd love to invite to dinner; but check out who is on the board from VMware:

  • Steve Herrod – Chief Technology Officer and Senior Vice President of R&D, VMware
  • Rod Johnson – General Manager, SpringSource; Senior Vice President of Middleware, VMware

Now Steve has been our CTO for quite a while and is instrumental in building the private cloud and steering VMware technology to be in alignment with the evolution of IT. But that second member is Rod Johnson, now the general manager of the SpringSource unit of VMware. Rod founded the Spring Framework, the most usable (and most used) Java framework today, and he is now charting out the course of building the most usable and powerful toolset for creating the next generation of cloud-enabled applications.

What does that say about building the private cloud, the platform-as-a-service stack that VMware is building, and the no-longer-even-interesting view of EMC as just a vendor of spinning rust? (Relevant link: Joe Tucci on the number of EMC software vs hardware engineers.) I leave that discussion up to you — it's just interesting is all I'm sayin' …

Reporting live from EMC World, this is Dr John Troyer for VMware

EMC World 2010 – let’s talk private cloud – live streaming

Journey to the private cloud[Executive summary: We'll be blogging about private cloud for the next three days here at EMC World 2010. Stuck at your desk? Follow along with EMC World 2010 live video from SiliconANGLE's #TheCube.]

I'm here in Boston to attend EMC World 2010. The theme, plastered everywhere at the venue and across the city (that's a sign at the taxi stand at Logan Airport over to the right) is "The Journey to the Private Cloud Starts Now." Thousands of infrastructure geeks are here to get their EMC on, and I'll be covering the more-virtualization-related activities at the conference.

I'm very interested in what people are thinking now about private cloud and if it's starting to make sense to them. Sometimes I see people have made it easy on themselves by just categorizing "cloud" as a marketing term for now, full of hype. That way they don't have to think about it too much. Others, coming from experiences with the public cloud, can't see how a private cloud makes sense. The reality is far more complex, and far more interesting. Already after talking to some of the attendees and EMC employees at the conference I'm seeing a recognition that private cloud concepts are resonating with the IT professionals. 

What makes private cloud challenging to talk about is that "private cloud" isn't a term the marketing gang cooked up that we can all ignore, like the normal Three Letter Acronyms that every company "is a leader in." Cloud is not a technology — you can't rack up a Cloud Box and call it a day. Cloud is an operational model, a consumption model, and way of running IT as a value-added part of your business. It's been fascinating so far this trip as everybody is closest to their part of the coalescing cloud — topics like provisioning, automation, chargeback, policy-based management and compliance were common topics at the opening reception — and it will be interesting to see how these threads come together over the next three days as we continue to talk public cloud.

EMC's Ed Saipetch has been briefing customers on the virtual datacenter and private cloud this week. He wrote yesterday about the conversations he's having. It seems to be a lightbulb type moment — to some IT shops, building a private cloud seems obvious, and to others, it still seems impossible. As Ed says in his post talking about some We’re separating but will stay friends:

Some customers saw exactly where I was going and others probably thought I was insane.  I started at a high level and then went into the details but here’s the problem. When we talk about infrastructure becoming a pool of resources that you’re able to push and pull workloads into and out of, some people think it’s fairytale land. … Infrastructure AND platforms are both part of the “Stack” and “Cloud” conversation.  It’s about businesses being able to let their most valuable asset (their people) work on deploying applications faster instead of provisioning servers.

We seem to be mirroring the same conversation we had when we had to give up being server huggers and virtualize in the first wave of server consolidation. As we're moving to a higher level of abstraction and automation, at first it seems like you're giving up what you've spent all your hard time to build in the first place. 

Sound like an interesting conversation? Let's get started. There's a few ways to follow along over the next few days.

And my favorite… streaming video LIVE from #TheCube at EMC World 2010. SiliconANGLE's John Furrier (@furrier) and Mark Rizzn Hopkins (@rizzn) will be broadcasting live Monday, Tuesday, and Wednesday from 2:30-5:30pm EDT (and a little more besides). They're set up right outside the bloggers lounge here, and they've already penciled in a great series of guests. You should not miss this show. Check it out at http://siliconangle.com/emcworld2010/. EMC's Stu Miniman (@stu) lays out more information on the broadcast and agenda:

What content are you getting?  THREE keynote speakers (2 of which will be coming directly from the stage straight to the live video), top industry bloggers & analysts, partners (VMware, Cisco, Brocade and others) and more special guests to be announced later.  Topics will span everything from Federation, Cloud and Virtualization to Social Media, Security and Sustainability.

Here's the current schedule:

On with the show! Reporting live from EMC World, this is Dr John Troyer for VMware!
John and Polly Pearson
John and EMC's @pollypearson after the Counting Crows on Sunday