Managing and Monitoring Data Collection > Virtualization Manager Data Collector Policy
  
Version 9.2.01
Virtualization Manager Data Collector Policy
VMware Data Collection offers two options for collecting data:
Collect data from vSphere (formerly vCenter) and/or
Collect data from specific ESX servers that are not managed by vSphere
When you create a Data Collection policy, you can specify host names and/or IP addresses or range of addresses. In addition, you can enter a comma-separated list of any combination of host names and addresses to be probed.
Prerequisites: A Data Collector must exist in the Portal, to which you will add Data Collector Policies. For specific prerequisites and supported configurations, see the APTARE StorageConsole Certified Configurations Guide.
 
 
 
Best Practice to Improve Performance
If you configure the vCenter 5-minute summaries for level 3 or higher, the Data Collector will use this configuration to reduce the load on the Portal during data collection. In vCenter, go to Administration > vCenter Server Settings > Statistics. Edit the 5-minute statistics interval and set the statistics to level 3. Note that when using level 3, some per-instance data (such as CPU core) may be missed or incomplete.
1. Select a Data Collector from the list.
2. Click Add and select VMware.
3. Configure the VMware Data Collector policy.
Field
Description
Sample Value
Collector Domain
The domain of the collector to which the collector backup policy is being added. This is a read-only field. By default, the domain for a new policy will be the same as the domain for the collector. This field is set when you add a collector.
 
Policy Domain
The Policy Domain is the domain of the policy that is being configured for the Data Collector. The Policy Domain must be set to the same value as the Collector Domain. The domain identifies the top level of your host group hierarchy. All newly discovered hosts are added to the root host group associated with the Policy Domain.
Typically, only one Policy Domain will be available in the drop-down list. If you are a Managed Services Provider, each of your customers will have a unique domain with its own host group hierarchy.
To find your Domain name select Admin > Hosts and Domains > Domains.
 
ESX/Virtual Center Server*
One or more IP addresses or host names to probe. Comma-separated addresses or IP ranges are supported. If the server requires an SSL certificate, refer to Adding a Certificate to the Java Keystore for instructions.
192.168.0.1-250, 192.167.1.10, myhost
User ID*
The view-only VMware user ID that has a role with the following privileges: Read-Only and Browse Datastore.
 
Password*
The password associated with the User ID.
 
ESX Servers to Exclude
Enter one or more ESX server names or IP addresses. Comma-separated names/addresses or IP ranges are supported.
 
Optimization Level
Select an optimization level for your VMWare data collection. Choose from default, aggressive, or none. Each selection impacts runtime differently. Effects will be most noticeable when running the Datastores collection.
 
Inventory
By default, this probe collects data from VMware using the defined schedule.
For performance reasons, this collection only gathers information on files currently in use by VMs collected by this probe. This includes VM disk files, suspended state, etc. Due to limitations in the data returned from VMWare, this probe may miss certain files, such as ISO files that are attached to virtual DVD drives.
To capture all datastore file information including the files associated with VMs in the inventory, select the check box: Datastore Scan. If the Datastore Scan probe is not regularly run, the Datastore Usage Breakdown report is likely to have a large Unknown category. For additional information, see Datastore Scan Collection and Reports Impacted by Datastore Collection.
Click the clock icon to create a schedule. Every Minute, Hourly, Daily, Weekly, and Monthly schedules may be created. Advanced use of native CRON strings is also available.
NOTE: Explicit schedules set for a Collector policy are relative to the time on the Collector server. Schedules with frequencies are relative to the time that the Data Collector was restarted.
*/30 * * * *
Datastore Scan
Select Datastore Scan to perform a complete scan and capture all datastore file information. This probe may take several hours to complete and should be configured to run when VMWare is less active (for example on weekends). This probe will find files that are not discoverable using the Inventory probe but are consuming space on the datastore. After this probe has run the Datastore Usage Breakdown report will normally only have a small Unknown category, however, there usually will be some Unknown data, representing file system overhead and internal VMWare disk structures that are not reported as files. Only enable this probe when actively looking for files that may be wasting space in the datastores. For additional information, see Datastore Scan Collection.
Click the clock icon to create a schedule. Every Minute, Hourly, Daily, Weekly, and Monthly schedules may be created. Advanced use of native CRON strings is also available.
NOTE: Explicit schedules set for a Collector policy are relative to the time on the Collector server. Schedules with frequencies are relative to the time that the Data Collector was restarted.
 
ESX Server Performance
Select ESX Server Performance to collect performance data for ESX servers.
Click the clock icon to create a schedule. Every Minute, Hourly, Daily, Weekly, and Monthly schedules may be created. Advanced use of native CRON strings is also available.
NOTE: Explicit schedules set for a Collector policy are relative to the time on the Collector server. Schedules with frequencies are relative to the time that the Data Collector was restarted.
 
Virtual Machine Performance
Select this probe to collect performance data from virtual machines. This probe may need to be run less often than the ESX Server Performance probe because there are usually more virtual machines than ESX servers. This probe may be disabled if it is not necessary to monitor the performance of individual virtual machines.
When the default schedule of once every 3 hours is used, the Data Collector will use vCenter’s 5-minute summaries for part of the data or, if collecting directly from an ESX server, there may be gaps in the data. To ensure that real-time data is always used (subject to the statistics level that is set), or if collecting from an ESX server, change the schedule of this probe to every hour (or more frequently). Click Help to view the statistics level settings in the Best Practice to Improve Performance.
Click the clock icon to create a schedule. Every Minute, Hourly, Daily, Weekly, and Monthly schedules may be created. Advanced use of native CRON strings is also available.
NOTE: Explicit schedules set for a Collector policy are relative to the time on the Collector server. Schedules with frequencies are relative to the time that the Data Collector was restarted.
 
Datastore Scan Collection
By default, Inventory collection only captures information about the files referenced by a collected virtual machine (VM files), and only using the information available without directly contacting the datastore. Examples of large files that will not be collected during the default Inventory collection are .iso files (such as, DVD images) and disk images (.vmdk) that are not attached to any collected VM. Data collection runtime is reduced when using the default Inventory collection, but only files for VMs in the inventory are collected. Therefore, a considerable amount of consumed space in the datastore may not be captured. Choose the Datastore Scan option to gain visibility into the storage that is not associated with a VM. See Reports Impacted by Datastore Collection.
When you check Datastore Scan, a more detailed collection is performed by scanning the datastore to discover files resident on the datastore, but not associated with a VM in the inventory. This option, although it may impact performance, finds files that may be invisible to VMware, but that are consuming space on the datastore. For example, a VM may have been removed without deleting files from the datastore. Similarly, log files or user-stored files may be taking up space not apparent from the VM inventory’s perspective. See Reports Impacted by Datastore Collection.
NOTE: If an error occurs during the Datastore Scan, then only the files available with the default Inventory collection will be collected for the Datastore. It is also possible that individual files may be found via the Inventory collection, but not via the Datastore Scan.
Reports Impacted by Datastore Collection
The following reports may be impacted by the Datastore Scan collection configuration in the data collector policy.
Report
Default Inventory
Collect all files on datastore
VM Files Summary (drilldown)
No Last Modified date.
Total # files collected will usually be less than with an all files collection.
Last Modified date will be missing for files not found in the datastore scan.
Datastore Usage Breakdown
Non-VM Files and VM Not In Inventory will be zero.
Total VM Used will be reduced by the amount that would have been collected as VM Not In Inventory if an all files collection had been done.
Unknown category in the pie chart will often be significantly higher than if an all files collection had been done.
The Unknown category in the pie chart will usually be non-zero and will often be significantly higher than the default inventory collection.
Datastore Utilization
VM Not In Inventory will be zero.
Total VM Used will be reduced by the amount that would have been collected as VM Not In Inventory if an all files collection had been done.
 
Datastore Detail
Non-VM Files and VM Not In Inventory will be zero.
Total VM Used will be reduced by the amount that would have been collected as VM Not In Inventory if an all files collection had been done.
VMDK and Other VM Files values typically will not be impacted by the datastore collection type.
VM Server Detail
VM Not In Inventory will be zero
Total VM Used will be reduced by the amount that would have been collected as VM Not In Inventory if an all files collection had been done.
Last Modified date will be missing for files not found in the datastore scan.
 
VM Detail
No Last Modified date.
Total # files collected will usually be less than with an all files collection.
 
VM Snapshot Summary
No Last Modified date.