Using the Instance Action List for Troubleshooting

If you need to troubleshoot an instance that has stopped working, you might find useful information about the issue in the Instance Action List.

The list shows every action you performed on the instance using the Compute service (Nova), such as creating, resizing, or rebooting it. By looking at the details of an action, you can determine whether it succeeded. Identifying a failed action can help you determine the cause of the failure. Viewing the sequence of actions that preceded the failure can also be helpful. 

Note: The list does not include actions performed with services other than Compute. For example, attachment to a volume would not appear on the list because that action involves the Block Storage (Cinder) service.

Using the Dashboard

  1. Choose INSTANCES from the PROJECTS drop-down list.
  2. In the INSTANCES table, click the name of the instance you want to troubleshoot.
  3. Click the Action Log tab.

    instance_action_list.png

  4. From the Instance Action List, click the request ID for any action that you want to view details for. The ID begins with req-.
  5. In the INSTANCE ACTION DETAIL table, view the result of the action to determine if it succeeded or failed.

    instance_action_detail.png

Using the Command Line Interface (CLI)

View a list of instances if you need the name or ID of the instance you want to troubleshoot:

$ openstack server list
+----------------+----------------+--------+----------------+---------------------------------------+
| ID             | Name           | Status | Networks       | Image Name                            |
+----------------+----------------+--------+----------------+---------------------------------------+
| <INSTANCE_ID>  | 0_snap         | ACTIVE | silver=<IP>    |                                       |
| <INSTANCE_ID>  | instance2      | ACTIVE | silver=<IP>    | CentOS-7-x86_64-GenericCloud-150_8-mc |
|                |                |        |                |                                       |
|                |                |        |                |                                       |
| <INSTANCE_ID>  | ceph_vol_snap_ | ACTIVE | silver=<IP>    |                                       |
|                | inst1          |        |                |                                       |
|                |                |        |                |                                       |
+----------------+----------------+--------+----------------+---------------------------------------+

View a list of all actions performed on the instance you want to troubleshoot:

$ nova instance-action-list instance2
+---------------+------------------------------------------+---------+----------------------------+
| Action        | Request_ID                               | Message | Start_Time                 |
+---------------+------------------------------------------+---------+----------------------------+
| create        | req-<UUID>                               | -       | 2017-03-06T21:29:36.000000 |
| resize        | req-<UUID>                               | -       | 2017-03-06T21:46:21.000000 |
| confirmResize | req-<UUID>                               | -       | 2017-03-06T21:51:55.000000 |
| reboot        | req-<UUID>                               | -       | 2017-03-06T21:58:10.000000 |
| stop          | req-<UUID>                               | -       | 2017-03-06T22:16:21.000000 |
| start         | req-<UUID>                               | -       | 2017-03-13T17:34:56.000000 |
+---------------+------------------------------------------+---------+----------------------------+

View the details of an action: 

$ nova instance-action instance2 
+---------------+----------------------------------------------------+
| Property      | Value                                              |
+---------------+----------------------------------------------------+
| action        | create                                             |
| events        | [{u'event': u'compute__do_build_and_run_instance', |
|               |   u'finish_time': u'2017-03-06T21:29:43.000000',   |
|               |   u'result': u'Success',                           |
|               |   u'start_time': u'2017-03-06T21:29:37.000000',    |
|               |   u'traceback': None}]                             |
| instance_uuid | <INSTANCE_UUID>                                    |
| message       | -                                                  |
| project_id    | <PROJECT_UUID>                                     |
| request_id    | req-<UUID>                                         |
| start_time    | 2017-03-06T21:29:36.000000                         |
| user_id       | <USER_UUID>                                        |
+---------------+----------------------------------------------------+

Tip: If you open a Metacloud Support request, you can expedite it by including the request ID for any failed action in the ticket. The Support team references these IDs in log files when troubleshooting.

Example: Using the List to Investigate a Stopped Instance

If an instance stops working unexpectedly, a quick way to investigate the reason is to view the details of the stop action in the list. One of the properties listed with the details is user_id.

If there is a value associated with user_id property, you know that a user initiated the stop. You can correlate the ID with the user, either in the Dashboard by clicking Users in the Admin drop-down list, or by running
openstack user show [USER_UUID]
You can then ask the user why he or she initiated the stop.

If there is no value associated with the user_id property in the stop action details, it's likely that the cause was beyond user control, for example, the failure of the hosting hypervisor. Make sure to indicate this to Support when initiating a request.

 

Have more questions? Submit a request
Powered by Zendesk