Introduction
Hello again, it's your friendly neighborhood Technical Marketing Engineer and this time I brought a few friends along for this blog. Long ago in a data center far away, an admin prayed the in-process upgrade would finish before the maintenance window expired. Hoping that something wasn’t missed in the long checklist the admin had gone over several times before hitting the enter key or clicked the mouse button in the UI to start the upgrade. Nutanix understands this operational challenge, which is why we have continued to design and refine the upgrade experience with operational simplicity in mind. Our latest AOS 7.6 release will surely give our admin friend - and the thousands more who rely on Nutanix - even more of the peace of mind and confidence that Nutanix is known for delivering.
For AOS 7.6, we evaluated updates intended to help reduce measured AOS upgrade elapsed time. To evaluate elapsed time under these conditions, our team clocked upgrade performance in our lab environment, evaluating both idle state and simulated workload conditions.
Lab Performance & Testing Data
To pull this blog together, we joined forces with the Nutanix Performance Engineering team. Because some customer environments may continue running workloads during maintenance, the team tested the upgrade with and without a running workload.
Our goal was to measure AOS upgrade elapsed time under these different test conditions. Now, I will caveat this by saying that with all things performance, your mileage may vary. These controlled lab results outlined below are not a guarantee of upgrade duration or performance in a customer environment. Upgrade duration can vary based on hardware, software versions, cluster configuration, workload, network conditions, operating procedures and environments, among other things. Ok, I know you are thinking enough talking, give me the details already.
Test Setup was as follows
- Hardware:
- Four NX-3060-G8 with Intel Xeon(R) Gold 5318Y CPUs
- 512GB Memory per node
- (6) 4TB SSDs each node
- Mellanox ConnectX-6 NICs
- Workload:
- Nutanix X-Ray simulated virtual desktop infrastructure (VDI) workload (for more info on Nutanix X-Ray head on over HERE)
- 252 virtual machines (63 per node)
- Each workload VM has (2) vCPUs, 4GB memory, a 100GB disk and a single NIC
- Each workload VM was running a mixed IO VDI workload
To put our updates to the test, we evaluated the upgrade paths under the following documented test conditions: a baseline upgrade with no active workloads and a test under load running the VDI workload noted above. We charted our flight path across two upgrade sequences: jumping from AOS 7.3 to 7.5, and then from AOS 7.5 to 7.6. Quick note, there wasn’t a 7.4 release and as such this is why you see the versions we used here.
Scope: Measured elapsed time refers to the AOS portion of the tested workflow, including AOS prechecks and postchecks. AHV and firmware update durations are excluded and not captured.
Wrapping Up: These Are the Upgrades You’re Looking For
Nutanix continues to refine infrastructure-update workflows designed to help simplify upgrade planning and execution. And never resting on our past successes, we took feedback to heart and went to work to improve. For additional information about related AOS 7.5 and AHV 11 updates, see the related article located HERE. In this controlled test environment and under the documented test conditions, the AOS 7.5 to 7.6 upgrade path produced a measured reduction in elapsed time compared with the AOS 7.3* to 7.5 path (* again, note there was no AOS 7.4) under each tested condition. Together, these measurements illustrate Nutanix’s ongoing focus on streamlining upgrade workflows.
As always, thank you for stopping by and taking a read!
©2026 Nutanix, Inc. All rights reserved. Nutanix, the Nutanix logo and all Nutanix product and service names mentioned are registered trademarks or trademarks of Nutanix, Inc. in the United States and other countries. All other brand names mentioned are for identification purposes only and may be the trademarks of their respective holder(s). This content reflects an experiment in a test environment. Results, benefits, savings, or other outcomes described depend on a variety of factors including use case, individual requirements, and operating environments, and this publication should not be construed as a promise or obligation to deliver specific outcomes.
