Jetstress Performance Testing on Nutanix Acropolis Hypervisor (AHV) – Part 1 – The Baseline Test

The following is Part 1 of the Jetstress performance testing on Nutanix Acropolis Hypervisor (AHV) series of videos.

This video shows the following:

  1. Stopping/Starting the NDSF cluster to ensure a fair starting point (No artificial pre warming of cache etc)
  2. The Performance required for 2500 Exchange Users (100 messages / Day with 2 DAG copies) being 732 Jetstress IOPS as per MS Exchange Server role requirements calculator.
  3. The Performance achieved by Jetstress with 8 threads using 8 vDisks (4 for DB, 4 for Logs)

The reason the demonstration is limited to 2500 users is because the Virtual machine compute requirements already is over the maximum recommended RAM for an Exchange 2013/2016 Server (96GB). As such, no additional storage performance is required as compute is more often than not the constraining factor.

For more information see: Peak performance vs Real World – Exchange

Note: This demonstration is not showing the peak performance which can be achieved by Jetstress on Nutanix. In fact it’s running on a ~3 year old NX-3450 with Ivy Bridge processors and Jetstress is tuned (as the video shows) to a low thread count which still achieves >3x the required IOPS for 2500 Exchange users.

Part 1

Return to the Table of Contents

Jetstress Performance Testing on Nutanix Acropolis Hypervisor (AHV) – Part 4 – Performance after three VM Migrations

This is Part 4 of the Jetstress performance testing on Nutanix Acropolis Hypervisor (AHV) series of videos.

This video shows the following:

  1. Jetstress performance on the forth node in the cluster after the VM has already been migrated in quick succession from its original node too a 2nd and 3rd node both of which have ran a Jetstress test.
  2. Nutanix Controller VM (CVM) CPU utilization

This test is the forth of several test to show the impact to performance by migrating VMs around several nodes in the same cluster which results in not all data being accessible/stored locally.

These post migration tests are designed to show the worst case performance when business critical workloads with large data sets (such as Exchange) move around nodes within a Nutanix AHV cluster. By running Jetstress on multiple nodes in quick succession, the workload on the Nutanix Controller VMs (CVMs) to localize data and maintain a even balance of data across all drives (SSD/HDD) and nodes within the cluster is higher than normal.

While migrating business critical applications around multiple nodes in quick succession (especially while running intensive benchmarks at much higher IOPS that the real world requires) is not something that will occur (unless initiated by an administrator), it demonstrates the performance impact (or lack thereof) of even this unrealistic operation.

Note: As with previous videos, This demonstration is not showing the peak performance which can be achieved by Jetstress on Nutanix.

Part 4

Return to the Table of Contents

Related Articles:

Jetstress Performance Testing on Nutanix Acropolis Hypervisor (AHV) – Part 3 – Performance after two VM Migrations

This is Part 3 of the Jetstress performance testing on Nutanix Acropolis Hypervisor (AHV) series of videos.

This video shows the following:

  1. Jetstress performance after the VM has been migrated in quick succession from its original node to a 2nd node, which runs a Jetstress test and then the performance of the VM on a third node within the same NDSF cluster.

This test is the second of several test to show the impact to performance by migrating VMs around several nodes in the same cluster which results in not all data being accessible/stored locally.

These post migration tests are designed to show the worst case performance when business critical workloads with large data sets (such as Exchange) move around nodes within a Nutanix AHV cluster. By running Jetstress on multiple nodes in quick succession, the workload on the Nutanix Controller VMs (CVMs) to localize data and maintain a even balance of data across all drives (SSD/HDD) and nodes within the cluster is higher than normal.

While migrating business critical applications around multiple nodes in quick succession (especially while running intensive benchmarks at much higher IOPS that the real world requires) is not something that will occur (unless initiated by an administrator), it demonstrates the performance impact (or lack thereof) of even this unrealistic operation.

Note: As with previous videos, This demonstration is not showing the peak performance which can be achieved by Jetstress on Nutanix.

Part 3

Return to the Table of Contents

Related Articles: