Installation (general)

Which versions of vCenter Server are supported by VMware Infrastructure Planner?
  • Versions 4.1 and above of vCenter Server are supported.
  • All vCenter Servers that will be connected need to be English language.
Are there limitations on the type of hosts that VMware Infrastructure Planner must be installed on?

The VMware Infrastructure Planner local collector appliance must be installed on a 64-bit host.

Where can I find detailed instructions for how to complete the VMware Infrastructure Planner installation and configuration?

A detailed installation guide is available here.

I had to enter the addresses of my vCenter Server, vRealize Operations Manager, and proxy servers in the configuration pages. Are these secured?

The collector appliance is hosted internally in your environment, and we also require the user to enter an assessment key before accessing the collector. This means that anyone accessing the collector appliance would need to be inside your network, know the IP address of the VMware Infrastructure Planner virtual appliance (which would likely require that they already have access to the vCenter Server), and know the unique assessment key to your assessment in order to access the collector appliance. Only hostnames and usernames are exposed to users with the assessment key; passwords are never shown to users. The collector appliance itself is secured with a random password, so no one can login.

When trying to configure the collector OVA the following message is displayed "Key is invalid. If you believe this is an error, please contact your sales engineer". What should I do?

This situation happens if you are trying to reuse an assessment key already connected with a collector. You may try to restart a collection or had to reinstall the OVA after having connected it without being able to complete the configuration. In order to recover from this situation, a new assessment needs to be created, the collector OVA needs to be re-installed and the new assessment key should be used.

Which types of proxies does the VMware Infrastructure Planner collector appliance support?

You may need to configure the collector appliance to use a web proxy communicate with the portal if your network does not allow direct access to the public internet. VMware Infrastructure Planner supports HTTP/HTTPS proxies. Basic and digest authentication are supported, but not NTLM, which is commonly used in Windows environments.

What is the minimum memory requirement for the VIP collector appliance?

The memory requirement for the VIP collector appliance is 8 GB.

vCloud Suite Assessments

Which versions of vRealize Operations Manager are supported by VMware Infrastructure Planner for vCloud Suite Assessments?
  • Versions 5.7.0 and above are supported. If customers already have vRealize Operations Manager, please ensure it's version 5.7.0 or above. Customers who don't already have vRealize Operations Manager can download a free version as part of, and for the duration of, the VMware Infrastructure Planner assessment.
  • The vRealize Operations Manager that will be connected needs to be English language.
I've correctly configured my VMware Infrastructure Planner virtual appliance. How long will I have to wait before I can view the results of a vCloud Suite assessment?

The length of time for data collection to complete depends primarily on whether your vRealize Operations Manager instance was installed as part of this VMware Infrastructure Planner assessment or was already installed.

If your vRealize Operations Manager instance was installed as part of this VMware Infrastructure Planner assessment, data collection will take at least 7 days. This is the minimum time recommended time for VMware Infrastructure Planner to build up a picture of resource use in your environment.

If your vRealize Operations Manager instance was installed prior to this assessment, but less than 7 days ago, data collection will begin as soon as 7 days of data are available.

Otherwise, your vRealize Operations Manager instance already has enough data, and data collection will begin immediately.

Data collection may take an additional 24 hours after collection begins.

You will receive an e-mail notification when the results are available to view online. Please reach out to your salesperson if you haven't received this notification after 24 hours if your vRealize Operations Manager was installed as part of this assessment, or after 8 days if your vRealize Operations Manager was installed prior to this assessment.

About Policies for vRealize Operations Manager

A policy is a set of rules that you define for vRealize Operations Manager to use to analyze and display information about the objects in your environment. How the policies are configured and applied and what threshold settings are set for those can affect the results found by VIP. So in order to get the most accurate results, please update the policies in vRealize Operations Manager appropriately to support the Service Level Agreements and business priorities established for your environment. Please refer to this documentation: VMware vRealize Operations Manager 6.0 Documentation Center to understand and update the policies.

The numbers I'm seeing in VMware Infrastructure Planner - for example, reclaimable vCPU - don't seem to match with the numbers reported in vRealize Operations Manager. Why is this?

There are a couple of reasons why this might be.

  • The first reason is that because we want to present a sensible view of the total savings opportunity across your environment, by default we enforce a policy that avoids double-counting. Specifically, if a VM is technically two or more of idle, powered-off, and oversized, then we enforce a hierarchy that places that VM in only one category. (See Is it possible for a VM to be eligible for both storage reclamation and vSAN? Doesn't this result in double-counting of storage savings? below for more on this). In contrast, a VM can appear in more than one of vRealize Operations Manager's idle VMs, powered-off VMs and oversized VMs views however, since the intention of vRealize Operations Manager is to identify all VMs in any of those categories, irrespective of how many categories a specific VM is in.

  • The second reason is that once VMware Infrastructure Planner has finished collecting data, it does not refresh the data further, while vRealize Operations Manager does.

What is the criteria for determining whether a VM is idle, powered-off, oversized or undersized?

We get data on idle, powered-off, oversized and undersized VMs direct from the most accurate data source we know - vRealize Operations Manager.

If you've installed vRealize Operations Manager as part of your VMware Infrastructure Planner assessment, we'll be identifying VMs based on the default vRealize Operations Manager policies. These are:

  • Idle VMs: VM is idle for at least 90% of the time analyze, where idle means the VM uses less than 100MHz CPU, 20KBps disk I/O, and 1 KBps network I/O
  • Powered-off VMs: VM is powered off for at least 90% of the time analyzed
  • Oversized VMs: CPU or memory demand falls below 30% for more than 1% of the time analyzed
  • Undersized VMs: CPU or memory demand rises above 70% for more than 1% of the time analyzed

If you already had vRealize Operations Manager, you may have adjusted these settings. You can review and adjust these settings from within vRealize Operations Manager.

Is it possible for a VM to be eligible for both storage reclamation and vSAN? Doesn't this result in double-counting of storage savings?

It is possible for a VM to be eligible for both storage reclamation as well as vSAN. A VM identified as "idle" for example will by definition be a candidate for reclamation; and since it'll also likely have recorded very low IOPS, it'll be identified as a good candidate workload for vSAN. Powered-off VMs and Snapshots are also candidates for reclamation, but we exclude them from vSAN cost savings total calculation, so that there is no double-counting.

VMware Infrastructure Planner is showing me fewer VMs than I have in my environment. Why?

There are a few reasons why this might be:

  • Did you add all your vCenter Servers to the VMware Infrastructure Planner virtual appliance configuration page? If you didn't, see below for instructions on how to add those.

  • Also on the VMware Infrastructure Planner virtual appliance configuration page, did you provide vCenter Server credentials that can see all VMs within the vCenter Server?

  • Has the number of VMs in your environment changed significantly since the VMware Infrastructure Planner assessment was completed? VMware Infrastructure Planner provides a snapshot of your assessment at a point in time, and won't update if your environment changes later. If this is important to you, ask your salesperson to create a new assessment. Then, access the existing VMware Infrastructure Planner virtual appliance configuration pages by navigating to the IP address of your VMware Infrastructure Planner appliance and appending /vcenterconfigurations/new, for example: On the final page of the configuration wizard, enter the assessment key from this new assessment.

The snapshots report seems to be counting all snapshots in my environment. Can I really take advantage of all these savings?

Right now, yes, the snapshots report does count all snapshots. We're working on features to make this more nuanced - for example, to only count snapshots that are outside the current VM branch. In the meantime, this number should be taken to represent a maximum possible saving; since you probably won't want to remove every snapshot - they're there for a reason after all! - you may want to mentally discount some portion of the snapshot savings.

How is disk IOPS in the Shared Storage report measured?

We collect one disk I/O measurement per hour, per VM, over a 7 day period: the average I/O recorded per second for that VM. This is the base data from which we calculate peak and average IOPS.

There are some parts of my environment that I don't want to include in any of the savings calculations - can I do this?

Yes - while a report is in draft mode, you'll need to ask your salesperson to perform these exclusions for you. You can exclude clusters in two different places; these two places are associated with different products, and therefore affect completely different reports.

  1. Click "Show Costs and Filters" in the top right corner of any page. This shows an assumptions bar that contains an area for excluded clusters. Clusters excluded here affect the output of reports associated with vRealize Operations and vRealize Automation, specifically the idle VMs, powered-off VMs, snapshots, oversized VMs and undersized VMs reports.

  2. From the Shared Storage report, click "Edit" below the "VMs excluded" label to search for VMs to exclude from the vSAN analysis only.

I downloaded the full report of vSAN-eligible VMs, but I don't see some VMs there. Why not?

For the Shared Storage report, we measure the IOPS requirements of all VMs contained in the vCenter Servers provided during configuration of the collector appliance.

There are a few reasons why VMs that you may expect to see are not included in the report:

  • Did you include the necessary vCenter Servers during configuration of the collector? If not, you will need to create a new assessment, download a new collector, and be sure to include that vCenter Server on the first step of the configuration wizard.

  • Did you or another member of the assessment manually exclude that VM from the Shared Storage report? You can check this from the Shared Storage report page by clicking on "Edit" below the "VMs excluded" label. Deselect any VMs that you don't wish to be excluded.

  • Has the VM reported higher IOPS per TB than the recommended vSAN threshold? vSAN is suitable for workloads with certain IOPS and read/write skew characteristics. As a proxy, for a typical VM, we suggest that VMs that record less than 2,000 IOPS per TB may be good candidates for vSAN. This figure can be changed via the Shared Storage report assumption. It could be that VMs you expect to see recorded IOPS data above this threshold.

  • Has the VM reported IOPS data at all? Since the Shared Storage report is based on analysis of IOPS data, if a VM does not report IOPS data, we won't include it at all in the analysis. Some of the reasons a VM may not report IOPS data is if the VM is stopped or paused.

I used the option to go to VMWare ROI TCO Calculator from the VIP assessment’s Settings tab and evaluate the ROI and TCO of using a specific VMware product/suite implementation based on data collected by VIP for this assessment. I find that that there are some discrepancies in the CAPEX Savings values that were indicated by VIP vs the ROI TCO Calculator. Why is that?

VMware Infrastructure Planner (VIP) performs an assessment of a customer’s virtualized environment and discovers savings. This assessment analyzes the existing vSphere virtual environment of the customer in order to identify potential resource savings arising from SDDC products, and their approximate value to customer. ROI TCO Calculator is a tool that helps estimate the overall economic value of implementing or upgrading your data center to the desired VMWare Product. There can be several reasons that can play in and cause a difference in the CAPEX value from the two tools. For example, in case of ROI TCO Calculator, the savings are those projected for the next 5 years vs immediate savings from reclaimable resources. Also, in case of ROI TCO Calculator the savings depend on the specific product (and version) you have and the one you are going to move to. There are some differences in the way firewall/networking related cost savings as well as storage related cost savings are calculated in the two tools. Another reason is that in VIP, the license costs are not included. If you see differences that cannot be explained by the documentation available for the individual products please contact VIP Support using the process described here.

vSAN Assessments

Which versions of ESXi are supported by VMware Infrastructure Planner for vSAN Assessments?

Versions 5.5 U1 and above are supported. VMs on ESX/ESXi hosts with a lower version number won't be analyzed by the collector appliance.

What network access do ESXi hosts require for a vSAN Assessment?

ESXi hosts that are running VMs to be analyzed use HTTP to download VIB from the collector appliance. Therefore, outgoing HTTP/TCP access on port 80 must be enabled. See the installation guide for detailed instructions.

Where should I deploy the collector appliance for a vSAN Assessment?

The collector appliance can be deployed on any host that has network access to the vCenter containing the VMs to analyze. The collector appliance must also be reachable by HTTPS (TCP port 443) by all hosts that are running the VMs to analyze, so that they may download the VIB from the collector appliance.

I've correctly configured my VMware Infrastructure Planner virtual appliance. How long will I have to wait before I can view the results of a vSAN Qualification assessment?

It takes about 2 days to collect the data needed for a vSAN assessment. It may take up to an additional day to send the data to VMware. You can check the status of your assessment by viewing it at

How does running a vSAN Assessment affect my environment?

So that the collector appliance can collect disk usage statistics, it installs a vSphere Installation Bundle (VIB) on each ESXi host that is running a VM selected for analysis. The VIB is an extension to ESXi that allows the communication of disk logging commands and data to and from the collector appliance, via the vSphere API. Installation of the VIB requires a restart of the hostd process, so heartbeats and other communication with the ESXi host may be momentarily interrupted. Once the VIB is installed, when disk usage is being collected, it may degrade disk I/O performance of workloads running on the host by 5 to 10%. When the assessment completes, the collector appliance stops collection of disk usage statistics, so performance should be restored to its original level. The VIB is currently not uninstalled after an assessment completes, but there should be no adverse effect of having it remain installed.

Is there a limit to the number of VMs or size of environment on which I can run the tool?

There is no limit per se, but there must be enough room on a datastore for each host to store tracefiles for all VMs being analyzed, in 10 minute chunks. The tracefiles contain a record of each I/O operation so for VMs that are performing many IOPS, the tracefiles can become large. As a rule of thumb, about 15 MB should be available for each VM being analyzed. The collector appliance will attempt to use the datastore with the most available space, but if not enough space is available, an error may occur. Also, the more VMs being analyzed, the more time it will take to complete collection.

How does vSAN work?

VMware vSAN works by replacing expensive SAN storage with cheaper network attached storage and SSDs on each host as disk caches. For good performance, most the data required at any time by a VM is stored on SSDs on the VM's host. For more details, see VMware's description of vSAN here.

What is the vSAN report showing me?

The vSAN report estimates the savings you would reap if you moved the VMs in the "Recommended vSAN cluster configuration" to a new cluster that uses vSAN, with the maximum cache size amount of SSD storage.

What is Histogram Analysis?

Histogram analysis takes summarized data about disk I/O operations and scores the suitability of a VM for vSAN from 0 (the least suitable) to 100 (the most suitable). Histogram analysis may not accurately determine if a VM is a good candidate for vSAN, but a low score does reliably indicate that a VM is unsuitable for vSAN. VMs with low scores are not analyzed further using tracefile analysis.

What is Tracefile Analysis?

Tracefile analysis takes a detailed log of disk I/O operations and computes the cache size required such that the disk I/O operations would achieve a high (by default 90%) hit rate if vSAN were used. The log of disk I/O operations only includes block addresses but not the actual contents being read or written. Tracefile analysis is more accurate than histogram analysis, but also causes a small performance degradation on the VMs it is measuring (about 5 to 10%).

What are the "I/O characteristics" that the vSAN report is referring to?

I/O characteristics, which affect the suitability of VMs for vSAN, include the following:

  • Whether the operations are reads or writes
  • The size of the blocks being read or written
  • The distance in bytes between the addresses being read or written (sequential versus random)
What is the active working set size?

The active working set is the data that a VM is actively reading or writing. If the active working set fits on an SSD on the VM's host, most I/O requests will be served locally with high performance. Therefore the size of the active working set determines how much SSD capacity is desired.

How can I change the assumptions for the vSAN cluster being built?

To change the assumptions for the vSAN cluster, select the link "Report Assumptions" at the top of the report. A dialog box will appear allowing you change the number of failures to tolerate, current SAN cost, SDD cost, HDD cost, maximum amount of cache for the cluster, and the required cache hit rate.

What is the maximum amount of cache for the cluster?

The maximum amount of cache for the cluster is the amount of SSD storage that will be installed on the hosts in the vSAN cluster being built. The SSD storage will act as a cache of the data stored somewhere on the network.

What is the number of failures to tolerate? How does it work?

Much like a RAID, vSAN can be configured to tolerate disk failures by storing multiple copies of data on different storage devices on the network. Of course, additional disk capacity is required for this capability. For each disk failure that is tolerated, the storage capacity required increases by 1X the amount of base capacity.

What cache hit rate should I use?

The default rate is 90%, which has been shown to produce good results. Increasing this number will result in higher performance at the cost of additional SSD storage. Conversely, decreasing this number will result in lower performance but less SSD storage required.

Why were some of my VMs unsuitable for vSAN?

Some reasons why a VM might be considered unsuitable for vSAN are:

  • The VM may perform too many IOPS that are not cachable, so its performance would be reduced significantly
  • The cache size required by the VM does not fit into the maximum amount allocated for SSD storage
  • The VM may have an access pattern that has causes the cache to be polluted – that is, filled with less cachable data that evicts cachable data that could have been used by other VMs
  • The VM may have been deleted while the assessment was running
  • The VM may have no virtual disks (VMDKs) associated with it
What is the difference between the vSAN Light Assessment carried out as part of the vCloud Suite Assessment and the vSAN Assessment?
  • vSAN Light Assessment carried out as part of the vCloud Suite Assessment is a very simplified analysis based on IOPS. If some VMs record low IOPS (below a settable threshold) they are identified as a good candidate for vSAN.
  • vSAN Assessment on the other hand is a much more detailed assessment done based on analyzing the details of the I/O operations. This assessment determines if a VM is a good candidate for vSAN or not and the required hard drive and SSD for the new (recommended) vSAN cluster.