Submit a ticketCall us

Webinar: Web Help Desk for HR, Facilities and Accounting Departments
This webinar will focus on use cases for HR, Facilities and Accounting.

Having a unified ticketing and asset management system for all the departments in your company can provide end-users with a seamless experience and make things easier for your IT team. Yet, with different business tasks and objectives, many departments don’t fully understand the capabilities of Web Help Desk and how the software can be customized for effective use in their departments.
Register Now.

Home > Success Center > Virtualization Manager (VMAN) > VMAN Collection Performance adjustments for better performance in large environments

VMAN Collection Performance adjustments for better performance in large environments

Table of contents
Created by Kevin Sperrazza, last modified by MindTouch on Jun 23, 2016

Views: 105 Votes: 1 Revisions: 9

Overview

This illustrates a few things that you can adjust on a VMAN master appliance to get a better performance from collections in very large virtual environments.

Environment

VMAN 6.X

Steps

  1. Get all collections off of the Master appliance and running on federated collectors.
  2. If you have multiple large VCenters, split them up between a few federated collectors so that you have about two data sources per collector.
  3. If the data source is smaller, you can easily fit more on a single collector.
  4. Once the collectors are in place, raise the memory limit of ActiveMQ so that the collectors are able to send the data back to the Master successfully in a large environment. This document shows what this issue looks like if you have it and how to adjust the memory settings in broker.properies file on the Master appliance.
  5. After the ActiveMQ has been adjusted, make some adjustments to the indexing. The changes you make here will be very low impact in what is missing from the data but large performance gains on the master appliance.
  6. Go to the VMAN admin console at https://vmanip:5480.
  7. Once you are logged in there under the Solarwinds Mgmt tab, you'll see an option to "Edit Configuration".
  8. From the drop down list chose the "hyper-config.xml" file and hit edit. This should bring up a text box with the configuration inside.
  9. Scroll to the bottom of the file and add the following two entries to the file.
    <entry key="index.lucene.search.lazyFetching">true</entry>
    <entry key="index.lucene.shallowIndex">true</entry>
  10. Once the file has been updated, click save and it should ask you at the top above the text box to restart the tomcat service for the changes to take effect. This is what it should look like:
  11. Once you hit the restart button at the top it will recycle the tomcat service on the appliance and the main Web Console will go down.
  12. Once the service has come back up the website at http://vmanip should return normally. When it's back up go ahead and login now as an admin. After you've logged in as an admin user we'll want to go to Setup->Advanced Setup->System Properties to make the last two changes for performance.
  13. On the System properties page search for the following settings and set them to "true"
    • datacollection.sample.ignore.basic.config.data 
    • This will skip changes in virtualization tree during performance collection. Drawback: New VMs or VMs moved to another host will be visible only after Configuration Job
    • collection.vmware.skip_storagepath_samples
    • This will skip collection of per-StoragePath samples which are not necessary for VMAN functionality, only per-StoragePath drill-down graphs will be missing

 

Shorten the retention period of raw data (Disk Size issues):

 

  1. In the Virtualization Manager appliance, click Setup > Advanced Setup > System Properties.
  2. Change the value of the Days to Retain Performance Data Hourly Rollups property to 30 (90 by default).
  3. Change the value of the Days to Retain Raw Performance Data property to 5 (14 by default). In this case, detailed graph data will not be available for the past days.


Delete old alerts and trends:

 

  1. In the Virtualization Manager appliance, click Setup > Advanced Setup > System Properties.
  2. Change the value of the months.to.keep.alerts property to 6 (12 by default).
  3. Change the value of the months.to.keep.trends property to 6 (12 by default).

 

No service restart is needed for these changes to take effect. Reboot all federated collectors at this time to clear out the jobs and get everything running with the new parameters in place.

 

Last modified
04:21, 23 Jun 2016

Tags

Classifications

Public