VMware vCenter chargeback is a product I use today. I use it for show back for internal IT costs. I am working to get it tuned to the point I can show where the money I spend in CapEx and OpEx is actually going. I think the holy grail of configuration for it would be balancing the spends vs usage seen, showing that each dollar we spend is either available or consumed resources. I also wish to use vCenter Chargeback to show that I provide a value back to the business. Cost transparency in how I am running the infrastructure is something i have tried to show, and that needs to exist going forward. Chargeback helps greatly with that.

From the session we can gather a few points.
1. A data collector is good for up to 15,000 VM’s or 5 vSphere instances.
2. Charge back managers scale out for performance and high availability.
3. Max config is around 35,000 VM’s and 10 vCenter servers.
4. I have heard via various conversations that people want API’s into chargeback. It looks like that is occurring. Also a SDK is to be released in Q4 2011

I encourage you to visit the chargeback community on VMware.com. I frequently browse them and there is some good info available there.

Bookmark to

For the past 2 years I have been trying to get more value out of vCO. It has seemed like vCO was a dead product. Today the room if nearly full with hundreds of people in this session. I am very happy to see this. I hope there is more content like plugins and workflow libraries available in them communities for vCO.

One of the first slides eludes to vCO being used to connect VMware Service Manager to vCloud. I hope this happens quickly, I really want this type of capability and agility. VCO isn’t about replacing the existing management systems. VCO is to be used to automate the manual tasks you perform in those management systems. Having VSM kick off workflows and automatically input the variables needed for the workflow to being is one of my goals at my company, happy it was mentioned here. I am very interested in VMware’s management strategy, I think we can expect good things.

One of the slides in this presentation shows VMware’s current it infrastructure from a high level. The next slide shows their end to end business application provisioning. I need to make sure to check these 2 slides out further.

Looks like there is a beta plugin for vCO to control F5 BigIP application delivery controllers. Application blueprints is a term we will be hearing more it seems. Finally it seems we are becoming aware that there is more to delivering an application that just a web server and a database.

VMware has seen 90%+ reduction in time for provisioning VM’s, applications, and scaling applications up with moving to cloud and using vCenter Orchestrator. Days and weeks to hours.

Plugins for vCO are being developed and seem to have some focus. Available on the VMware solutions portal. Hopefully a community can be created around these plugins and hopefully workflows that can help customers get a good jumpstart on using these tools. A plugin sdk and certification program is being created. Single stop market place for plugins is currently in beta.

There is a new instructor course for developing workflows. There will be self paced online training coming late 2011. There is also a book “Automating vCenter 5.0 with vCO” to be released in 2011. There is a blog for vCO with examples and best practices. This is actually where i saw the UCS plugin in action.

Hopefully VMware develops a virtual appliance for all of their software. A vCO appliance is in the works.

Well, let’s get started reducing our OpEx by using vCenter Orchestrator.

TEX2923 and TEX1442 are two additional sessions to look at for vCO.

Bookmark to

Lab Manager does a great job of providing environments on the fly via self service. vCloud Director appears to be positioned to provide the same service in a better way.

One of the first slides in this session shows the terminology map between vCloud director and lab manager. I myself need to remember to go and look at this session online and get a copy of that slide.

Lab manager only supports 1 vCenter, but vCloud director supports 25 vCenters today. VCloud director 1.5 supports SQL now which is a nice change.

This session was mostly about SAP’s massive lab manager deployment and PoC of vCloud director to be the successor of this. They have their work cut out for them for migrating all of their VM’s in lab manager to vCloud.

Bookmark to

I recently acquired vCenter Operations enterprise. I have had a PoC done by VMware to ensure it met my requirements of a consolidated view of what is really going on in my converged infrastructure. Although I have some extensive knowledge into vCOps I am very interested in getting an under the covers look at how some of the data is calculated. I am also interested on how best to monitor my operation and what KPI’s I should really be looking at. Then I want to learn how best to interpret some of them. Let’s see what this session has to offer….

1. Lol first slide says it’s not just black magic! I like this session already.
2. With virtualization capacity is now fluid, I agree.
3. Invisible walls, with vm CPU and memory issues may not be resolved by adding more, contention can play a significant role. Proper troubleshooting is required.
4. With vmview you have to monitor end users not VM’s. I agree here as well as a user may move between virtual desktops. End user experience is important.
5. The key thing for VC OPs to do for me is to take the tons of metrics I have and to present the end calculation to me. Am I green or red?
6. Dynamic threshold analysis uses competing algorithms, meaning the system actually uses multiple methods to calculate the trend, then checks to see who is right more often and then uses that method. Genius. These are calculated every night.
7. A version change can cause the normal operation of a system to change. Thresholds may not catch this but trends have a better chance. I liked this idea.
8. Trending noise to determine abnormalities, this is really going to help my environment since we use a number of tools that are all sending emails for every little thing. We use our brains today to get a feel for the data center health. I declared this a broken model earlier in the year.
9. Alerts should be an indication of a real problem, yes yes yes! Do not alert on every threshold that is reached. Yes please. May I subscribe to your newsletter!
10. Root cause determination in this product is really root metric determination. It isn’t telling you what the problem was just what metric that was being monitored was the starting metric of the issue. I.e. We saw disk latency go to 100ms before the app crashed.
11. Workload is demand divided by entitlement.
12. Right-sizing is a concept I always support, but VM admins and I seem to be on the front line of this alone on this.

Bookmark to

I like vCenter Configuration Manager more and more as I use it. The reporting and control it gives me is much better than my previous product. I am pointing at my Altiris server with the dust on it. Who am i kidding, I virtualized that product when we installed it. The thought of a physical server not running esx on it is laughable to me. I am going to sit here and eat my pretzel while we wait for the session to start. I assure you that your wait will be shorter than mine.

See? Let’s begin.

1. This session was a little weird because the presenters would interrupt each other and start talking. Odd.
2. VCM integration with vCloud Director looks cool. Auto discovery and understanding of machines with the same name and ip but in different zones.
3. The change detection is tricky for me. I want to monitor everything that could possibly change because of paranoia.
4. Although I do not use it, VCM can provision OS, via pxe boot.
5. Someone agrees that the cloud should include moving the workload back and forth between the public and private cloud.
6. One thing I had not really considered before but I am now, is giving read only access and reports to the server app owners in my org. We keep this info to the server guys right now.

Bookmark to

1. The iterative memory copy that vMotion does was interesting to hear more about in detail. I think too many people may not quite understand what a vMotion is really doing. I often just explain it as black magic.
2. VSphere 5 interests me, the idea of multiple nics being used for vMotion. I wonder how best to configure this in UCS. Is vMotion traffic more than client traffic?
3. VMotion ID once found can give details about vMotion in the vmkernel log.
Tests appear to show that vMotion on vSphere 5 is faster, 37% drop in time on vSphere 5.
4. vMotion obviously gains tremendous performance on 10GbE.
5. Multiple vMotion NICs and vMotion performance gain in vSphere 5.
6. Configure vMotion NIC’s on the same vSwitch.
7. For best vMotion performance place vm swap on SAN.
8. If you use CPU reservations, leave some CPU for the host for vMotion ~30%.
9. Metro vMotion supported in vSphere 5. Doesn’t help us VPlex Geo guys.
10. Conclusion? Upgrade to vSphere 5 ASAP if you want faster vmotions.
11. I’m headed to ask about vMotion over asynchronous distances. If not NDA I will share

Good session for vMotion people.

Bookmark to

Automation and IT Service Management are big interest points for me this year. VMAXe and VPlex integrations and best practices also are in scope for me this year. To top it all off I plan to spend as much time in the hands on labs as possible.

If you see me out there feel free to chat me up on topics of interest. I still need to find out if there is a bloggers lounge and take care of some last minute scheduling.

See you there!

Bookmark to

In an effort to quickly determine how bad VMWare’s vSphere 5′s new licensing is going to hit the company wallet I wrote a powershell one liner. It needs vsphere powershell to function. Once connected to vCenter the first line will show the sum of configured memory in MegaBytes.

> get-vm | measure-object -property MemoryMB -sum
Count : 518
Average :
Sum : 2171984
Maximum :
Minimum :
Property : MemoryMB

Since VMWare’s new licensing says that an Enterprise Plus Processor license is good for up to 48 GB you can do some quick math with the numbers to see how many licenses you need. 2171984 / 1024 = 2121.078125 GB. 2121.078125 GB of configured memory is just over 2 TB. 2121.078125 / 48 GB = 44.18 licenses. So it appears I need 45 processor licenses to run vSphere 5 with my current memory configurations.

This second line I have added a few more switches. These will show you your largest, smallest, and average memory configuration. The average is useful so that you can figure out on average how much a VM costs you from this new licensing view point.

> get-vm | measure-object -property MemoryMB -sum -average -maximum -minimum
Count : 518
Average : 4193.0193050193
Sum : 2171984
Maximum : 32768
Minimum : 256
Property : MemoryMB

VMWare vSphere Enterprise Plus retails for $3495 and provides you with licensing for 48 GB of memory. The means each GB you configure on a VM roughly costs you $72.81. My average VM using the numbers above costs $298.14 just for the vSphere 5 license.

This new model will prove interesting in the near future as better understanding of it is developed. It seems to me that management will see that a VM has yet another license cost. This one tied directly to how much memory we configure and usually waste on a VM. Memory increases on a VM that doesn’t consume that memory now have a cost associated with them.

This new model seems like it will hurt some of the large memory configuration blade providers that have 2 or 4 procs but support 256 or 512 GB of memory. Their prices just went up from $6,990 and $13,980 to $18,639 and $37279 My Cisco B200-M2 blades will continue to cost me $6,990 since they have 96GB of memory. I am concerned about those new B440 series Cisco blades I was looking at. 4 Procs but 512 GB of memory, $37,279 in vsphere licenses per box? Ouch.

All prices are list. There are other editions of vSphere that are cheaper however I do not use those. vSphere 5 does have a lot of positve new features that I am unsure if I can live with out. Storage DRS for instance is something we have been asking for for years.

Bookmark to

EMC World has ended and I am on a plane thinking about Ionix UIM. This is a product that appears to be geared towards configuration of vBlock. I’m interested in answering some questions for myself. I am curious if I can automate the provisioning of new blades as they arrive at my location. I am also interested in automating the provisioning of new chassis as they are plugged into my UCS fabric interconnects.i am interested in automating the full provisioning and configuration of ESXi. Tis could help me move a long the path of getting out of the server provisioning business.

Currently provisioning a new blade is extremely easy with UCS. I said easy not instant, fast, or man power free. Unpack and slide in sounds great and not worrying after this point sounds great. Minutes to a couple of hours later have VM’s start vMotioning to it would be awesome.

Today we have to log into UCS and assign a service profile. Then we zone it in the brocade switches and assign a boot LUN in unisphere. Now we boot it and mount an iso to install ESXi. Then we add the rest of the storage to it that is our normal datastores. Assign a host profile that never seems to be perfect and configure the network. Near this point some final tweaks and we have VM’s on it. A process that seems to take us a couple of days at best.

I do not think UIM is going to support brocade but it might. In he help file I saw only mention of MDS. I bet it does support nexus though. I’ll get this questions answered soon enough. The lab the product seem simple from a functionality point of view but it did meet a key piece of strategy to remove manual process and replace with automation. It also appeared to fill a gap in physical capacity reporting in a single pane though not as good as I want. I would love to see integration with capacity iq to show true capacity including virtual capacity.

All in all UIM is something I plan to look at further.

Bookmark to

EMC World 2011 is well underway here in Las Vegas at the Venetian convention center.

I will be posting notes and information I am able to gather and share from the sessions and speakers.

Bookmark to