Problem with delete and revert snapshots via API rest request

[Amdusis]

Good day!

We can successfuly create snapshot on VM

{
  "action" : "create",
  "resource": {
    "name" : "test01"
  }
}

{
  "results": [
    {
      "success": true,
      "message": "Creating snapshot test01 for Virtual Machine id:1000000032923 name:''",
      "task_id": "1000000021171",
      "task_href": "https://fqdn/api/tasks/1000000021171"
    }
  ]
}

But cant delete or revert to snapshot, miq answer thats all "OK" but tasks not even reach vCenter
Delete:

{
  "action" : "delete"
}

{
  "success": true,
  "message": "Deleting snapshot test01 for Virtual Machine id:1000000032923 name:''",
  "task_id": "1000000021173",
  "task_href": "https://fqdn/api/tasks/1000000021173"
}

In UI we can see that task state is "Active" but nothing doing

In logs we can see the followng

INFO -- api: MIQ(Api::VmsController.log_request) API Request:    {:requested_at=>"2025-09-08 10:10:53 UTC", :method=>"POST", :url=>"https://fqdn/api/vms/1000000032923/snapshots/1000002495810", :request_id=>"ff57c209-3157-44d9-9991-5a8ffe5873d7"}
INFO -- api: MIQ(Api::VmsController.log_request) Request:        {:method=>:post, :action=>"delete", :fullpath=>"/api/vms/1000000032923/snapshots/1000002495810", :url=>"https://fqdn/api/vms/1000000032923/snapshots/1000002495810", :base=>"https://fqdn", :path=>"/api/vms/1000000032923/snapshots/1000002495810", :prefix=>"/api", :version=>"4.4.0-pre", :api_prefix=>"https://fqdn/api", :collection=>"vms", :c_suffix=>nil, :collection_id=>"1000000032923", :subcollection=>"snapshots", :subcollection_id=>"1000002495810", :request_id=>"ff57c209-3157-44d9-9991-5a8ffe5873d7"}
INFO -- evm: MIQ(MiqQueue.put) Message id: [1000142292150], Zone: [default], Role: [], Server: [], MiqTask id: [1000000021174], Handler id: [], Ident: [generic], Target id: [], Instance id: [1000000032923], Task id: [], Command: [ManageIQ::Providers::Vmware::InfraManager::Vm.remove_snapshot], Timeout: [600], Priority: [100], State: [ready], Deliver On: [], Data: [], Args: [1000002495810]
INFO -- evm: MIQ(MiqTask.generic_action_with_callback) Task: [1000000021174] Queued the action: [Deleting snapshot test01 for Virtual Machine id:1000000032923 name:'s-msk-t-miq-api03'] being run for user: []
INFO -- evm: MIQ(MiqQueue.get) Message id: [1000142292150], Zone: [default], Role: [], Server: [], MiqTask id: [1000000021174], Handler id: [MiqWorker 1000000117339], Ident: [generic], Target id: [], Instance id: [1000000032923], Task id: [], Command: [ManageIQ::Providers::Vmware::InfraManager::Vm.remove_snapshot], Timeout: [600], Priority: [100], State: [dequeue], Deliver On: [], Data: [], Args: [1000002495810], Dequeued in: [1.834393308] seconds
INFO -- evm: MIQ(ManageIQ::Providers::Vmware::InfraManager::Vm#raw_remove_snapshot) removing snapshot ID: [1000002495810] uid_ems: [2025-09-08T10:01:47.066402Z] ems_ref: [snapshot-2479193] name: [test01] description []
INFO -- evm: MIQ(ManageIQ::Providers::Vmware::InfraManager::Vm#run_command_via_parent) Invoking [vm_remove_snapshot] through EMS: []

In vCenter logs or tasks nothing about request from manageiq.

vCenter version is 8.0.3 build 24853646
ManageIQ version - Radjabov

Any help please) thanks!

[Fryguy]

@agrare Can you take a look at this one? Can't tell if it's a bug or perhaps we're just not timing out properly?

[agrare]

I think the issue is the API queues the snapshot operations for the wrong queue,

 #<MiqQueue:0x00007f86837957a0
  id: 5498,
  target_id: nil,
  priority: 100,
  method_name: "remove_snapshot",
  state: "ready",
  created_on: "2025-09-17 18:32:10.973718000 +0000",
  updated_on: "2025-09-17 18:32:10.973718000 +0000",
  lock_version: 0,
  task_id: nil,
  deliver_on: nil,
  queue_name: "generic",
  class_name: "ManageIQ::Providers::Vmware::InfraManager::Vm",
  instance_id: 281,
  args: [2],
  miq_callback: {:class_name=>"MiqTask", :instance_id=>24, :method_name=>:queue_callback, :args=>["Finished"]},
  msg_data: nil,
  zone: "default",
  role: nil,
  server_guid: nil,
  msg_timeout: 600,
  handler_id: nil,
  handler_type: nil,
  expires_on: nil,
  tracking_label: nil,
  user_id: nil,
  group_id: nil,
  tenant_id: nil,
  miq_task_id: 24>]

That queue_name: "generic" should be queue_name: "ems_operations"

It worked for me initially because I was using simulate_queue_worker as a quick test and that dequeues for all queue_names but yes I think this would fail with full workers.

[agrare]

@Amdusis I opened a PR to fix the API, thanks for the bug report! ManageIQ/manageiq-api#1298

[Amdusis]

@agrare @Fryguy
Big thanks! All is works now!