Today, I am highlighting a new service in vSphere 7: ESXi Health Perspectives, or simply put, EHP. This is a service for checking if a given host or cluster can be transitioned to a given target state part of its lifecycle. For example, other services can call EHP to determine if a host can safely be removed from the cluster.
Part of these basic lifecycle tasks are:
- Can the ESXi be safely put into maintenance mode?
- Can the ESXi be safely removed from the cluster?
- Is the ESXi ready to receive VM/workloads?
EHP increases the reliability of ESXi health status reports, prior to tasks for managing the lifecycle of ESXi hosts.
Integration with vCenter Server Services
EHP is linked to vCenter. I mean it cannot run independant of vCenter as no EHP agent exists on the ESXi. EHP is installed on the vCenter Appliance. No additionnal steps required to configure it.
ESXi Health Perspectives calls to other services to evaluate the state of the host. These services called Health Providers, used by EHP, are:
- vpxd
- vSphere HA, DRS
- vSphere ESX Agent Manager (EAM)
- vSAN
- and more on the roadmap …
So, currently there is no agent on the ESXi, EHP is running on the vCSA. Actually, EHP is triggered for any ESXi/cluster lifecycle operation that is managed by Update Coordinator/Personality Manager.
Integration with Update Coordinator
The only caller of EHP is the Update Coordinator
The Update Coordinator is a subcomponent of Personality Manager.
Personality Manager drives the lifecycle operations on the ESXi (upgrades/reconfig.)
EHP is hosted behind the Update Manager vAPI endpoint.
Architecture of ESXi Health Perpectives
When the EHP is called by Update Coordinator, before performing any particular lifecycle operation, EHP calls to its health providers ( VPXP, DRS, vSAN as depicted below) to see if there are any issues: depicted as “Evaluate Health”.
EHP aggregates the identified issues to provide a OK/NOT_OK (or WAIT) status. Furthermore, EHP passes this information collected from the health providers, vCenter will determine the good health state to continue or not the safe lifecycle operation.
On the whole, EHP consists of a set of objects responsible for acquiring and aggregating ESXi health status information.
For example, EHP will check that vSAN has evacuated all of its data from the host.
Reporting Health issue
If an issue is reported, vSphere Client presents the error/warning.
The reported issues provide very clear messages, self-explanatory, to help you investigate further in logfiles, for example, and resolve the issue prior to attempting the task (upgrade or whatever).
Troubleshooting EHP: the log file
The EHP log vmware-vum-server.log is located in:
/storage/log/vmware/wmware-updatemgr/vum-server/
Search in the log file for “Successfully set up EHP” to make sure EHP has started correctly.
When investigating in the log file, use the keywords “ESX Health Perspectives“, as per example: 
Check our Latest posts
- Bassem Rezkalla Discusses NSX ALB on VMware CMTY Podcast
- 🇱🇺 Luxembourg to Host Its First Ever VMUG!
- vSphere 8.0 update 1a – Build 21815093
- The traditional vSphere 8 update 1: “What’s new?”
- MysticMarvin: The Blog Strikes Back
- ChatGPT is not the only AI!
- The VMware Entire Portfolio in 1-click – February 2022
- Denis’s Trek For Lisa: a Phenomenal Sporting and Solidarity Challenge
- vSAN Datastore: Storage Fault has occurred
- VMworld 2021 Final CountDown!
- Marvin’s notes from the field: VxRail ACE Lost Connection
- vSAN 7 Troubleshooting Tools
