Back to all use cases

ConsultingCross-Industry

Plan the move before the support date and check the failure headroom on the way

Engineering Organization

A typical consulting engagement for a team planning for vSphere 8 end of general support on 11 October 2027: a compatibility check, an upgrade order with a pilot cluster, and a review of whether each cluster survives a host failure.

VMware consulting

Overview

A typical engagement starts with an infrastructure team that runs vSphere 8 and knows its end of general support is 11 October 2027. The team has no written plan, is not sure its hardware, backup tool and monitoring agents are ready, and has never checked what a host failure would do to its clusters.

Challenge

An upgrade touches more than ESXi: vCenter has to be at or above the hosts' version, and hardware, backup products, monitoring agents and storage drivers each have their own compatibility lists. Clusters sized years ago may also have no spare capacity: when a host fails, vSphere HA restarts its virtual machines on the remaining hosts only if admission control has reserved room for them, and a cluster that has drifted can leave some virtual machines powered off.

Environment

Several vSphere 8 clusters on mixed server generations, vSAN on some of them, a backup product and monitoring agents integrated with vCenter, and a maintenance window of a few hours a month.

Approach

AceMQ engineers list every host, driver, firmware and integrated product and check each against its vendor's compatibility list. They pick a low-risk cluster as the pilot, write the order of vCenter and host upgrades, and plan a rollback for each step. For each cluster they compute what is left after the largest host fails and compare it with the admission control policy, and read CPU ready and co-stop to find oversized virtual machines that waste capacity.

Solution

  1. Inventory of hosts, firmware, drivers, vCenter and integrated products
  2. Compatibility check of each item against its vendor's list
  3. Upgrade order: vCenter first, then a pilot cluster, then the rest, with a rollback per step
  4. Failure headroom calculated per cluster and compared with the HA admission control policy
  5. CPU ready and co-stop read in esxtop to find oversized virtual machines
  6. A schedule that fits the monthly maintenance window

Outcome

The engagement delivers a compatibility register, an upgrade order with a pilot and rollbacks, a headroom figure for each cluster, a list of virtual machines to right-size and a runbook for host loss. The team runs the upgrades itself, or AceMQ engineers run the pilot cluster beside it. Hardware vendor faults and Broadcom patches stay with those vendors.

Technologies

VMware vSphereVMware vCenterVMware vSAN

Need VMware vSphere Architecture Guidance?

AceMQ's senior VMware vSphere engineers have handled this exact type of engagement before. Whether you need architectural guidance, hands-on remediation, or an ongoing managed partnership, we're ready to help.