Azure Linux VM stuck in Updating; extension and managed Run Command stuck in Deleting

Question

Azure Linux VM stuck in Updating; extension and managed Run Command stuck in Deleting

Erfani Pak 0

Azure diagnostics confirmed a backend concurrency conflict on a production Linux VM. The VM is still running and the Azure VM Agent reports Ready, but the VM provisioning state remains Updating.

Resource:

Subscription: fb887a60-6d84-41c2-be31-27d1db3dbcf1
Resource group: CALYXA-PROD-CORE-RG
VM: CALYXA-PROD-APP-VM01
Region: UAE North
Resource ID: /subscriptions/fb887a60-6d84-41c2-be31-27d1db3dbcf1/resourceGroups/CALYXA-PROD-CORE-RG/providers/Microsoft.Compute/virtualMachines/CALYXA-PROD-APP-VM01

Current state:

VM power state: Running
Azure VM Agent: Ready
VM provisioning state: Updating
Extension enablevmAccess: Deleting
Managed Run Command calyxa-recovery-diag-20260818-001: Deleting

Azure Operation Failure Diagnostics detected:

Operation: VirtualMachines_RunCommand_POST
Time: 2026-08-18 11:55:54 UTC
Error: "run command extension execution is in progress. please wait for completion before invoking a run command."
Root cause reported by diagnostics: conflict between concurrent requests.

A VM Reapply completed successfully at 2026-08-18 13:01:20 UTC.

Reapply correlation ID: 506d2af4-a7a6-4293-bf83-a4896de59ed5

Six other stuck managed Run Commands were successfully deleted, but the two child resources above remain stuck. No further Run Command, restart, reapply, redeploy, resize, or delete operation is being issued.

Request:

Please escalate this Priority Community Support issue to a private Microsoft Compute support case and perform backend cleanup to complete or cancel the two stuck delete operations. The VM, OS disk, data disks, NIC, identity, networking, backup data, and application data must be preserved. This is not a VM-size availability issue.

Prrudram-MSFT 28,506 Reputation points Moderator

2026-08-24T08:32:15.8933333+00:00
Hello **Erfani Pak,

**Based on the details provided, this does appear consistent with a control-plane/backend operation lock caused by a concurrent Run Command operation, where the VM remains functional but the resource is blocked in an Updating provisioning state due to stuck child-resource delete operations. The symptoms match a backend operation conflict scenario requiring Compute engineering intervention rather than customer-side remediation

Please investigate and perform below remediation to:

Identify any orphaned or non-terminal ARM/Compute operations associated with the VM.

Cancel, complete, or clean up the following stuck delete operations:

Extension: enablevmAccess

Managed Run Command: calyxa-recovery-diag-20260818-001

Clear any stale Compute RP/ARM operation locks causing the VM provisioning state to remain Updating.

Return the VM resource to a terminal provisioning state (Succeeded).

Preservation Requirements:

This is a production workload.

Please preserve:

VM resource

OS disk

All data disks

NICs

Managed Identity

Networking configuration

Backup configuration/data

Application data

Important:

We do not recommend VM deletion or recreation.

Do not perform redeployments that risk workload impact unless explicitly coordinated.

This is not an Azure capacity or VM-size allocation issue.

Request is for Compute backend cleanup of stuck operations and recovery of provisioning state.

Reference document for provisional states of a VM: https://learn.microsoft.com/en-us/azure/virtual-machines/states-billing#provisioning-states

If issue persists, please revert to me with details requested via private message.

1 answer

Your answer

Prrudram-MSFT 28,506 Reputation points Moderator

2026-08-24T08:32:15.8933333+00:00

Hello **Erfani Pak,

**Based on the details provided, this does appear consistent with a control-plane/backend operation lock caused by a concurrent Run Command operation, where the VM remains functional but the resource is blocked in an Updating provisioning state due to stuck child-resource delete operations. The symptoms match a backend operation conflict scenario requiring Compute engineering intervention rather than customer-side remediation

Please investigate and perform below remediation to:

Identify any orphaned or non-terminal ARM/Compute operations associated with the VM.

Cancel, complete, or clean up the following stuck delete operations:

Extension: enablevmAccess

Managed Run Command: calyxa-recovery-diag-20260818-001

Clear any stale Compute RP/ARM operation locks causing the VM provisioning state to remain Updating.

Return the VM resource to a terminal provisioning state (Succeeded).

Preservation Requirements:

This is a production workload.

Please preserve:

VM resource

OS disk

All data disks

NICs

Managed Identity

Networking configuration

Backup configuration/data

Application data

Important:

We do not recommend VM deletion or recreation.

Do not perform redeployments that risk workload impact unless explicitly coordinated.

This is not an Azure capacity or VM-size allocation issue.

Request is for Compute backend cleanup of stuck operations and recovery of provisioning state.

Reference document for provisional states of a VM: https://learn.microsoft.com/en-us/azure/virtual-machines/states-billing#provisioning-states

If issue persists, please revert to me with details requested via private message.

Answer 1

Hello @Erfani Pak

You can't fix this from client side. It's a backend ARM lock.

State you have:

VM: CALYXA-PROD-APP-VM01 - Running / Agent Ready / ProvisioningState: Updating
Stuck child resources in Deleting: enablevmAccess + RunCommand calyxa-recovery-diag-20260818-001
Root cause: Concurrent RunCommand_POST at 2026-08-18 11:55 UTC - run command extension execution is in progress
Reapply 506d2af4-a7a6-4293-bf83-a4896de59ed5 succeeded at 13:01 UTC but didn't clear the lock.

What Microsoft has to do: Backend cancellation / ARM cache refresh of the 2 stuck delete operations for resource ID /subscriptions/fb887a60-.../CALYXA-PROD-APP-VM01. This is a known Compute RP concurrency bug - requires internal Compute support - ProvisioningState stuck escalation. No restart/redeploy/resize/delete should be attempted until cleared.

Open ticket as:

Type: Technical > Virtual Machine > VM provisioning state
Severity: A - Production down risk
Title: Backend cleanup needed - VM stuck Updating, extension + RunCommand stuck Deleting - concurrency conflict
Paste your Resource ID, correlation IDs, and state: Preserve VM, OS disk, data disks, NIC, identity, backup - no delete/redeploy.

If this answers your query, do click Accept Answer and Up-Vote for the same. And, if you have any further query do let us know.