Error: "Request processing is restricted by administrator" in performance charts tab in vCenter 7.x/8.x

Products

• VMware vCenter Server

Issue/Introduction

• When opening the performance charts for a cluster in vSphere Client (<cluser> → Monitor → Performance → Overview) the tab remains empty with an error message stating:
  
  Request processing is restricted by administrator.
  
  
• The performance charts tabs for other objects (resource pools, hosts, virtual machines etc.) show the same error.
• The vCenter service log, /var/log/vmware/vpxd/vpxd-<number>.log contains entries similar to:

  ####-##-## error vpxd[90632] [Originator@6876 sub=MoPerfManager opID=########] The query size of ### metrics exceeded the vpxd.stats.maxQueryMetrics limit of 256 metrics. Dropping

Environment

• VMware vCenter Server 7.0.x
• VMware vCenter Server 8.0.x

Cause

Any vCenter version from 5.5 Update 2d on, has a pre-configured limit to the number of entities that are included in a single database query.
This limit is meant to protect the vCenter Server database from receiving large queries, which might adversely impact vCenter performance.
If the number entities in the query exceeds the default limit, this issue occurs.

Resolution

To prevent this issue from occurring, the limitation for performance charts can be increased. This is a 2-Step process:

1. Edit the config.vpxd.stats.maxQueryMetrics vCenter advanced option.
2. Edit the web.xml configuration file of the web server used to render the charts.

Edit the config.vpxd.stats.maxQueryMetrics vCenter advanced option.

1. Open vSphere Client in a web browser and login with an account that has administrative permissions for the vCenter.
2. Select the vCenter object in the inventory tree and open Configure → Settings → Advanced Settings.
3. Click on Edit Settings
4. Search the list for "config.vpxd.stats.maxQueryMetrics".
   1. If the option exists, change the value next to it to the desired limit, then click on Save.
   2. If the value does not exist, enter config.vpxd.stats.maxQueryMetrics in the Name field below the table and use the Value field to provide the desired query size limit. Then click Add to add the advanced setting, before selecting Save.

Note: The default value for config.vpxd.stats.maxQueryMetrics is 256. Review the messages in the vpxd logs to evaluate the required value.
Alternatively using -1 as value will disable the limit.

Edit the web.xml configuration file of the web server used to render the charts.

1. Open an SSH session with the vCenter Server Appliance and login as root.
2. Navigate to /usr/lib/vmware-perfcharts/tc-instance/webapps/statsreport/WEB-INF:

   # cd /usr/lib/vmware-perfcharts/tc-instance/webapps/statsreport/WEB-INF

3. Create a backup copy of the web.xml:

   # cp /usr/lib/vmware-perfcharts/tc-instance/webapps/statsreport/WEB-INF/web.xml /usr/lib/vmware-perfcharts/tc-instance/webapps/statsreport/WEB-INF/web.xml_301449

4. Edit the file using the VI editor
5. Find the following block:

   <context-param>
      <description>Specify the maximum query size (number of metrics)for a single report. Non-positive values are ignored.</description>
      <param-name>maxQuerySize</param-name>
      <param-value>###</param-value>
   </context-param>

6. Review the value in the line starting with <value-param>. If it exceeds the value used in part 1, no further changes are needed.
7. Otherwise, change the value to the limit configured in part 1. Alternatively a value 0 will disable the limit.
8. Save the file and exit the editor (:wq!)
9. Restart the VMware Performance Charts service:

   # service-control --restart vmware-perfcharts

   Refer to Stopping, Starting or Restarting VMware vCenter Server Appliance services for more information.

Attention: Increasing the value of config.vpxd.stats.maxQueryMetrics beyond the default value of 256, or disabling the limit completely, will result in larger queries against the vCenter Database. This will put increased strain of the involved components and might result in:

• Deterioration in the general responsiveness of the vCenter Server
• Longer response times to display some of the performance charts

As a rule of thumb, use a value slightly above the query sizes reported in the vpxd logs.
For example, if the message in the vpxd log says this:

The query size of 277 metrics exceeded the vpxd.stats.maxQueryMetrics limit of 256 metrics. Dropping.

Select a value only slightly larger than 277, e.g. 280.

Additional Notes:

• Cluster metrics are dependent on host or VM metrics.
• A single cluster metric value might depend on a single or multiple host- or VM metrics.
• The maxQueryMetrics limit is exceeded when the metric query count exceeds the default value (256).