Please provide your feedback in this short Flings' survey.
fling logo of DRS Dump Insight

DRS Dump Insight

version 2.1 — November 30, 2021

Comments 59

  • profile picture of Anil Bhargav Paindla
  • profile picture of Shimona Verma
  • profile picture of Gabrie
  • profile picture of Anil Bhargav Paindla
  • profile picture of Gabrie
  • profile picture of Anil Bhargav Paindla
  • profile picture of Anil Bhargav Paindla
  • profile picture of Gopi
View All

Summary

DRS Dump Insight is a service portal where users can upload drmdump files and it provides a summary of the DRS run, with a breakup of all the possible moves along with the changes in ESX hosts resource consumption before and after DRS run.

Users can get answers to questions like:

  • Why did DRS make a certain recommendation?
  • Why is DRS not making any recommendations to balance my cluster?
  • What recommendations did DRS drop due to cost/benefit analysis?
  • Can I get all the recommendations made by DRS?

Once the drmdump files are uploaded, the portal provides a vMotion timeline based on the files uploaded, users can navigate through multiple dump analysis with individual summary (for each drmdump file) of all the possible vMotion choices DRS went through before coming up with the final recommendations.

vCenter version 6.0-6.7:

DRS dumps the cluster state every 5 minutes, this includes the host states and vm states along with DRS rules and options set in the cluster.

An example timeline & summary can be seen for versions 6.0-6.7(Figure 1)

The portal also enables users to do What-If analysis on their DRS clusters with options like:

  • Changing DRS Migration Threshold
  • Dropping affinity/anti-affinity rules in the cluster
  • Changing DRS advanced options
  • Figure 2 & 3 are sample screen shots showing What-If? analysis with migration threshold changed from 2 to 5

vCenter version 7.0 and above:

With the introduction of DRS 2.0 to vSphere 7.0, the dump format and the core algorithm have changed.

Each drmdump(Plmt-*.log) file contains many dumps categorized into two types:

  • Partial dumps: These dumps are created whenever a VM is powered on or migrated to another host.
    • Supported scenarios: Load balance, Initial placement, Maintenance mode and Mandatory moves (moves to correct rule violations)
  • Full dumps: These dumps contains information about all the active VM's and ESXI hosts present in a cluster, dumped every 3 minutes.

An example timeline & summary can be seen for versions 7.0+(Figure 4)

Users can get more information about the different types of moves considered, by clicking on the "Show" link next to the type of move that they're interested in. An example is shown in Figure 5.

Requirements

  • Supports drmdump files from vCenter versions 6.0 and above
  • The recommended browser is Google Chrome

Instructions

1. Upload a folder of dumps

  • Folder must contain unzipped dump files
  • You can find the drmdump files in the vCenter server appliance under the following directory: "/var/log/vmware/vpxd/drmdump/[cluster domain]/" (click on the cluster in the VC UI and locate the [domain-ID] in the url from the browser)

  • version 6.0-6.7:
    • The dump files are archived by default and named as: " *-proposeNonResActions.dump.gz "
    For version 7.0 and above:
    • The dump files are named as "Plmt*.log*"
  • Extract the dump files to be uploaded into a folder and upload the folder in the portal
  • Do not modify dump names or format else it will throw an error.

2. Select the appropriate vCenter version and click upload

  • For 7.0 and above, enable the toggle "All full dumps" to analyse all the full dumps.
  • Note: Leave the option in default state unless absolutely necessary, the operation is very resource intensive and can lead to significantly longer wait times.
version 60-67:
  • Each point on timeline represents a dump, clicking on a point shows the respective dump summary
  • Users can zoom( mouse scroll ), pan( mouse drag ) and reset( mouse right click ) the timeline view.
version 70 and above:
  • Each blue point on the timeline represents a full dump, each green point represents a partial dump.
    • A yellow point color indicates, DRS simulation failed (partial success, information like reason for moves is not available)
    • A red point indicates total analysis failure.
  • Clicking on a point shows the respective dump summary
  • Users can select an area by dragging the mouse to zoom into and reset( mouse right click ) the timeline view (Fig 6).


3. To export multiple dump analysis as pdf, select all the points corresponding to the dumps on the timeline( In the export pdf dialog ) and export.

Changelog

Version 1.1

  • Users can now upload multiple dumps as a folder.
  • Creates a vMotion timeline based on the dumps uploaded, users can navigate through multiple dump analysis.
  • Users can export multiple dump analysis as a PDF at once.
  • Added support to 65u2, 65u3 and 67u3 dumps.
  • Bug fixes and backend improvements

Version 2.0

  • Added support for 7.0 and 7.0U1 dumps.
  • Toggle added for selective analysis of all full dumps.
  • Bug fixes and backend improvements

Version 2.1

  • Added support for 7.0U2 and 7.0U3 dumps.
  • Bug fixes and UX/Backend improvements”

Similar Flings

No similar flings found. Check these out instead...
Aug 15, 2016
Jul 13, 2018
fling logo of vAssist.ai NLP Platform

vAssist.ai NLP Platform

version 1.0.0

vAssist.ai, a Natural Language Processing (NLP) platform, enables bot developers to train machine learning models for intent classification and entity extraction.

Apr 21, 2015
fling logo of PowerCLI Cmdlet for NFS

PowerCLI Cmdlet for NFS

version 1.0

VMware PowerCLI is one of the most successful command line tools for managing your VMware products. With many existing cmdlets designed for the system administrator or vSphere Admin, PowerCLI is the easiest and most powerful tool for managing your environment.

Nov 30, 2016
fling logo of ESXi Mac Learning dvFilter

ESXi Mac Learning dvFilter

version 2.0

This ESX extension adds functionality to ESX to support MAC-learning on vswitch ports. For most ESX use cases, MAC learning is not required as ESX knows exactly which MAC address will be used by a VM.

Sep 07, 2021
fling logo of Forklift for Workspace ONE UEM

Forklift for Workspace ONE UEM

version 3.0.2

The Forklift for Workspace ONE UEM Fling allows migration of Workspace ONE UEM Resources between different Workspace One UEM environments. Rapid deployment of resources bundled as a "Deployment Template". Lastly the creation of a Continuous Delivery Pipeline for resources.

Oct 14, 2020
fling logo of vSphere Pod Autoscaler

vSphere Pod Autoscaler

version 1.0

This Fling is useful for vSphere PodVM users who want to perform auto-scaling on vSphere PodVMs based on memory utilization.

View More