vRA 7.6 Looping Provision / Looping VMDispose

Mindwatering Incorporated

Author: Tripp W Black

Created: 09/15 at 12:36 PM

 

Category:
VMWare
vRA

Issue:
VM Provisions or destroy fail with event subscriptions looping after completion.


Cause 1:
If VM was migrated (exported w/metadata) from old vCenter and imported to another vCenter, and
if the original is not removed, and
if the original ESXi host is in maintenance mode blocking automation, and
if vRA and "see" both, and
if they have same UUID, then vRA's proxy agents will deliver the VMDispose to both the new and the old VMs.

This will fail on the old VM, and vRA will loop waiting on the VM to come out of maintenance mode.

Solution:
Delete the old VM.


Cause 2:
If the VM is on a single ESXi host or a cluster w/all hosts in maintenance mode for upgrades, then vRA will loop waiting on the VM to come out of maintenance mode.

Solution:
N/A. Since upgrades are typically quick. vRA will proceed once a host is out of maintenance mode that hosts the VM.


Cause 3:
The proxy agents or vRA looses it's mind.


Workarounds:
1. Cancel all jobs, open SR to have VMware clean the database. Try again.

2. VMware suggested one of the following to assist vRA to break the loop:
vRA --> Infrastructure (tab) --> Administration (menu twistie) --> Global Settings (menu option) --> Group:Agent (twistie right) -->
Edit the Errors subgroup document, Unrecoverable DisposeVM Errors.

Name: Unrecoverable DisposeVM Errors
Description: Unrecoverable errors returned from DisposeVM in Regex format
Value: ^Value cannot be null

or

Name: Unrecoverable DisposeVM Errors
Description: Unrecoverable errors returned from DisposeVM in Regex format
Value: ^Object reference not set to an instance of an object


VMware KB Articles:
2144906 - Request to destroy a machine fails in VMware vRealize Automation 7.0.x
2114385 - Deployment / Provisioning requests are stuck at "In Progress" or "Pending Approval" status


Notes:
VMware wanted more than the normal VAMI log bundle.
They wanted the postgres tables (follow KB 2074214) for vRA.
They wanted the IaaS database via the MS SQL Config. Mgr. program to create full backup of the db.
Be very careful updating postgres tables if deleting columns. Make sure you have a COLD snapshot first of the vRA environment VMs.

















Keywords:
VMDispose Destroy Looping Issue

previous page