[AMBARI-19055] Removing Tasks From host_role_command Causes Upgrades To Show As PENDING

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 2.2.0
• Fix Version/s: 2.5.0
• Component/s: ambari-server
• Assignee: Jonathan Hurley
• Reporter: Jonathan Hurley
• Created: 01/Dec/16 20:48
• Updated: 07/Dec/16 18:26
• Resolved: 07/Dec/16 18:26

Description

It may be necessary to remove entries from the host_role_command table if the size of the table has grown excessively large in order to reduce the query times for "IN_PROGRESS" requests.

However, if you remove all tasks for an upgrade, but you leave the stages, request, and upgrade items, the upgrade will appear as though it is now PENDING.

Before Removal

{
  "href": "http://localhost:8080/api/v1/clusters/c1/upgrades/12",
  "Upgrade": {
    "cluster_name": "c1",
    "create_time": 1480517560897,
    "direction": "UPGRADE",
    "downgrade_allowed": true,
    "end_time": 1480517643350,
    "exclusive": false,
    "from_version": "2.5.0.0-1237",
    "pack": "nonrolling-upgrade-2.5",
    "progress_percent": 0,
    "request_context": "Upgrading to 2.5.2.0-67",
    "request_id": 12,
    "request_status": "COMPLETED",
    "skip_failures": false,
    "skip_service_check_failures": false,
    "start_time": 1480517560950,
    "suspended": false,
    "to_version": "2.5.2.0-67",
    "type": "INTERNAL_REQUEST",
    "upgrade_type": "NON_ROLLING"
  },
  "upgrade_groups": [
  ...

After Removal

{
  "href": "http://localhost:8080/api/v1/clusters/c1/upgrades/12",
  "Upgrade": {
    "cluster_name": "c1",
    "create_time": 1480517560897,
    "direction": "UPGRADE",
    "downgrade_allowed": true,
    "end_time": 1480517643350,
    "exclusive": false,
    "from_version": "2.5.0.0-1237",
    "pack": "nonrolling-upgrade-2.5",
    "progress_percent": 0,
    "request_context": "Upgrading to 2.5.2.0-67",
    "request_id": 12,
    "request_status": "PENDING",
    "skip_failures": false,
    "skip_service_check_failures": false,
    "start_time": 1480517560950,
    "suspended": false,
    "to_version": "2.5.2.0-67",
    "type": "INTERNAL_REQUEST",
    "upgrade_type": "NON_ROLLING"
  },
  "upgrade_groups": [
  ...

The actual request is showing as COMPLETED correctly:

{
  "href": "http://localhost:8080/api/v1/clusters/c1/requests/12",
  "Requests": {
    "aborted_task_count": 0,
    "cluster_name": "c1",
    "completed_task_count": 0,
    "create_time": 1480517560897,
    "end_time": 1480517643350,
    "exclusive": false,
    "failed_task_count": 0,
    "id": 12,
    "inputs": null,
    "operation_level": null,
    "progress_percent": 100,
    "queued_task_count": 0,
    "request_context": "Upgrading to 2.5.2.0-67",
    "request_schedule": null,
    "request_status": "COMPLETED",
    "resource_filters": [],
    "start_time": 1480517560950,
    "task_count": 0,
    "timed_out_task_count": 0,
    "type": "INTERNAL_REQUEST"
  },
  "tasks": [],
  ...

STR:

• Perform a simple upgrade of a ZK-only cluster
• Find out the ID of the upgrade request (say its 12)
• Remove tasks

  DELETE FROM execution_command WHERE task_id IN (SELECT task_id FROM host_role_command WHERE request_id = 12)
  
  DELETE FROM host_role_command WHERE request_id = 12
  

Issue Links

• links to: Reviewboard