Events¶
An Event is a record generated during engine execution. Such an event captures what has happened inside the senlin-engine. The senlin-engine service generates event records when it is performing some actions or checking policies.
An event has a level
property which can be interpreted as the severity
level value of the event:
- 10: interpreted as
DEBUG
level. Events at this level can be ignored safely by users. For developers they may provide some useful information for debugging the code. - 20: interpreted as
INFO
level. Events at this level are mostly about notifying that some operations have been successfully performed. - 30: interpreted as
WARNING
level. Events at this level are used to signal some unhealthy status or anomalies detected by the engine. These events should be monitored and checked when operating a cluster. - 40: interpreted as
ERROR
level. Events at this level signifies some failures in engine operations. These event should be monitored and checked when operating a cluster. Usually some user intervention is expected to recover a cluster from this status. - 50: interpreted as
CRITICAL
level. Events at this level are about serious problems encountered by the engine. The engine service may have run into some bugs. User intervention is required to do a recovery.
Senlin provides an open architecture for event dispatching. Two of the
built-in dispatchers are database
and message
. The database
dispatcher dumps the events into database tables and it is enabled by default.
The message
dispatcher converts the event objects into versioned event
notifications and published on the global message queue. This dispatcher is
by default disabled. To enable it, you can add the following line to the
[DEFAULT]
section of the senlin.conf
file and then restart the service
engine:
event_dispatchers = message
Based on your deployment settings, you may need to add the following lines to
the senlin.conf
file as well. This lines set messaging
as the default
driver used by the oslo.messaging
package:
[oslo_messaging_notifications]
driver = messaging
Note that unprocessed event notifications which are not associated with a
TTL (time to live) value by default will remain queued at the message bus,
please make sure the Senlin event notifications will be subscribed and
processed by some services before enabling the message
dispatcher.
Since the event dispatchers are designed as plug-ins, you can develop your own event dispatchers and have senlin engine load them on startup. For more details on developing and plugging in your own event dispatchers, please refer to the Plugin Writer’s Guide document.
The following sections are about examining events when using the database
dispatcher which creates database records when events happen.
Listing Events¶
The following command lists the events by the Senlin engine:
$ openstack cluster event list
+----------+---------------------+---------------+----------+----------------------------+-----------------------+-----------+-------+
| id | generated_at | obj_type | obj_id | obj_name | action | status | level |
+----------+---------------------+---------------+----------+----------------------------+-----------------------+-----------+-------+
| 1f72eb5e | 2015-12-17T15:41:48 | NODE | 427e64f3 | node-7171861e-002 | update | ACTIVE | 20 |
| 20b8eb9a | 2015-12-17T15:41:49 | NODE | 6da22a49 | node-7171861e-001 | update | ACTIVE | 20 |
| 23721815 | 2015-12-17T15:42:51 | NODEACTION | 5e9a9d3d | node_delete_3e91023e | NODE_DELETE | START | 20 |
| 54f9eae4 | 2015-12-17T15:41:36 | CLUSTERACTION | 1bffa11d | cluster_create_7171861e | CLUSTER_CREATE | SUCCEEDED | 20 |
| 7e30df62 | 2015-12-17T15:42:51 | CLUSTERACTION | d3cef701 | cluster_delete_64048b01 | CLUSTER_DELETE | START | 20 |
| bf51f23c | 2015-12-17T15:41:54 | CLUSTERACTION | d4dbbcea | cluster_scale_out_7171861e | CLUSTER_SCALE_OUT | START | 20 |
| c58063e9 | 2015-12-17T15:42:51 | NODEACTION | b2292bb1 | node_delete_59da99f0 | NODE_DELETE | START | 20 |
| ca7d30c6 | 2015-12-17T15:41:38 | CLUSTERACTION | 0be70b0f | attach_policy_7171861e | CLUSTER_ATTACH_POLICY | START | 20 |
| cfe5d0d7 | 2015-12-17T15:42:51 | CLUSTERACTION | 42cf5baa | cluster_delete_352e1b6b | CLUSTER_DELETE | START | 20 |
| fe2fc810 | 2015-12-17T15:41:49 | CLUSTERACTION | 0be70b0f | attach_policy_7171861e | CLUSTER_ATTACH_POLICY | SUCCEEDED | 20 |
+----------+---------------------+---------------+----------+----------------------------+-----------------------+-----------+-------+
The openstack cluster event list command line supports various options when listing the events.
Sorting the List¶
You can specify the sorting keys and sorting direction when list events,
using the option --sort
. The --sort
option accepts a
string of format key1[:dir1],key2[:dir2],key3[:dir3]
, where the keys used
are event properties and the dirs can be one of asc
and desc
. When
omitted, Senlin sorts a given key using asc
as the default direction.
For example, the following command sorts the events using the timestamp
property in descending order:
$ openstack cluster event list --sort timestamp:desc
When sorting the list of events, you can use one of timestamp
, level
,
otype
, oname
, user
, action
and status
.
Filtering the List¶
You can filter the list of events using the --filters`
. For example,
the following command filters the event list by the otype
property:
$ openstack cluster event list --filters otype=NODE
The option --filters
accepts a list of key-value pairs separated by
semicolon (;
), where each pair is expected to be of format key=val
.
The valid keys for filtering include oname
, otype
, oid
,
cluster_id
, action
, level
or any combination of them.
Paginating the Query results¶
In case you have a huge collection of events (which is highly likely the case),
you can limit the number of events returned using the option
--limit
. For example:
$ openstack cluster event list --limit 10
Another option you can specify is the ID of an event after which you want to
see the returned list starts. In other words, you don’t want to see those
events with IDs that is or come before the one you specify. You can use the
option --marker
for this purpose. For example:
$ openstack cluster event list --limit 20 \
--marker 2959122e-11c7-4e82-b12f-f49dc5dac270
At most 20 action records will be returned in this example and its UUID comes after the one specified from the command line.
Showing Details of an Event¶
You can use the senlin command line to show the details about an event you are interested in. When specifying the identity of the event, you can use its name, its ID or its “short ID” . Senlin API and engine will verify if the identifier you specified can uniquely identify an event. An error message will be returned if there is no event matching the identifier or if more than one event matching it.
An example is shown below:
$ openstack cluster event show 19ba155a
+---------------+--------------------------------------+
| Field | Value |
+---------------+--------------------------------------+
| action | delete |
| cluster_id | ce85d842-aa2a-4d83-965c-2cab5133aedc |
| generated_at | 2015-12-17T15:43:26+00:00 |
| id | 19ba155a-d327-490f-aa0f-589f67194b2c |
| level | 20 |
| location | None |
| name | None |
| obj_id | cd9f519a-5589-4cbf-8a74-03b12fd9436c |
| obj_name | node-ce85d842-003 |
| obj_type | NODE |
| project_id | 42d9e9663331431f97b75e25136307ff |
| status | DELETING |
| status_reason | Deletion in progress |
| timestamp | 2015-12-17T15:43:26 |
| user_id | 5e5bf8027826429c96af157f68dc9072 |
+---------------+--------------------------------------+