I have noticed many partners lately complaining about their Capacity Planning report having below expected VM per server Density, & has normally been granted access to their Capacity Planning Project to optimize the report. I have decided I will share the info on here on which parameters to change in order to  get a decent VM per Server Density in your CP report. Although I have been using these numbers successfully During my job for a while, and I know that many of my colleagues are following the same I provide no explicit/implicit guarantee of the results nor my employer. If you think your capacity planner report got out of hand on results (Ex 1000 vm per host), then please check what you have done wrong or contact your local VMware SE, or feel free to leave me a comment below. I will repeat use at your own risk, & ensure to use your judgement before submitting the report to customers.

Bumping up CP consolidation ratios Procedures:

1- First Make sure you are using Consolidation Analysis (CA) rather than Consolidation Estimate (CE) before you follow up any of the below procedures, else you will never find out the required tabs. You can always choose to run Consolidation Estimate (CE) when first creating the CP project in VMware Capacity Planning, but what if you already have a project running as Consolidation Estimate (CE) then you can follow the below procedure to convert it to Consolidation Analysis (CA).

a- From the Setup Page in your Consolidation Estimate click on Edit Project Info as shown in the Screen Snapshot below

VMware Capacity Planning Dashboard Setup Page

b- In the Capacity Planning Dashboard Edit Project Info Page choose Consolidation Analysis (CA) & Hit OK as illustrated in the below screen snapshot

Change Consolidation Estimate(CE) to Consolidation Analysis (CA)

Now that you have Switched to Consolidation Analysis (CA), you should be able to change the require parameters for a better consolidation ratio report as below. Before you follow the step below make sure you have chosen the Company & the project under projects in the top left site of the CP dashboard.

1- It seems the default set target system thresholds for paging rate and file cache are set quite low, and for a better consolidation ration you should bump them up as below:

paging rate  200 page faults/sec  ==Change to==> 10,000 pages/seconds

File Cache 600MB             ==Change to==> 2147MB

Below are the steps to make the changes and get updated reports:

a. Go to Analyze->Optimization Scenarios and edit each of the 2 builtin scenarios (they both start with ***)

Choose Optimization Scenarios from the Analyze menu

b. Go through the wizard and edit the numbers on the Systems Thresholds screen (see attached screenshot)

c. Force the scenarios to be run again (click the checkbox next to the scenario and hit Run)

d. Rerun the optimization reports

This makes sense according to how ESX manages memory – each VM has its own paging file, and its own paging threshold in CP. So, the logic behind having a consolidated paging rate doesn’t make much sense.

When Editing your scenarios you will be presented with Specify Consolidation Rules page. On this page make sure you choose “Merge all architectures” as well all the options under “Cross Boundaries” as shown in the screen shot below, unless you have a special reason to do otherwise.

Merge Architectures & Cross Boundaries Snap Shot

Further recommended values for all of the parameters are:

· Processors: CPU Usage % – 60 (same as default)

· Processors: Queue – 4 (same as default)

· Memory: RAM Usage % – 90 (same as default)

· Memory: Page File Usage % – 100 (default is 70%)

· Memory: Paging – 10000 (default is 200)

· Memory: File Cache – 2147 (default is 600)

· Disk: Speed – 50 (same as default)

· Disk: I/O Speed – 5000 (default is 1000)

· Network: Speed – 1000 (default is 10)

I hope this help, & please leave your feedback in comments.

I know many Virtualization Consultants who all are still wondering where is the Intel Nehalem processor in the VMware Capacity Planner. The Nehalem processor has already hit the market more than 6 months back, VMware still has not update their Capacity Planner to include it. We all know that the Nehalem processor is providing a great performance & efficiency enhancement comparing to the prior Intel processors & this change should be taken into the consideration When doing Capacity Planning.

Customers has been as well wondering why the capacity planner report is made with an older processor, as they will be buying the Nehalem one. The only way I had found around this is to build your custom hardware inside the VMware Capacity Planner with a faster CPU frequence, but even this still not too accurate. I am hoping VMware will address this soon. If any one has any updates or VMware plans for this topic & would share it with us. it would be of a great value to everyone.

Offline VMware Capacity Planner:

I have mostly done online VMware Capacity Planner as most customers usually don’t mind it, though when I started offering VMware Capacity Planner for Banks & Ministries that’s when Offline VMware Capacity Planner came handy. Many of the companies & ministries mind you sending data online directly, where they are OK with you sending it after they inspect it. As VMware Capacity Planner allow you to queue all your inventory data & performance data on the Capacity Planner system and upload it from another system or at another time that problem can be over come by what I like to call Offline VMware Capacity Planner. Although you will need to use the same methodology if the system running the capacity planner does not have an internet & you want to upload the data at a later time using another machine with internet access. The procedure of Offline VMware Capactiy Planner is below as documented by VMware KB#1896:

Follow these steps:

1. Register the Database ID from the primary (originating) collector. The secondary (sending) collector does not need to be registered to the Company ID. For more information see: Viewing and Adding Collector Database IDs to a Client Registration (1891) .

2. On the originating Collector computer export the inventory data into a .CSV file. First setup the task to export only and process any Linux inventory data.  Click Admin > Options > Modules > Data Synchronize. Delselect all items except Export collection data for systems, and Import Collection Data for Systems.

3. Click Tasks > Run Manual Tasks > Run Data Synchronization Task. Wait for the process to finish.

4. Using Windows Explorer open the Capacity Planner base directory, typically %ProgramFiles%\VMware\VMware Capacity Planner .  Open the outbox folder.

5. Copy the files from outbox folder on the originating Collector to the outbox folder on the sending Collector. For each file copied, verify that the copy is intact before you delete that file from the source computer.

6. From the sending Collector upload the files copied into its outbox folder. Click Tasks > Run Scheduled Jobs > Run Scheduled – Data Synchronize. The data is now be securely uploaded from the sending Collector to the IW.

I hope that help some one find the instruction faster, as it took me some time to find them out few months back. Enjoy the tip & please leave me a comment if that help some one.