[jira] [Created] (KNOX-2410) CM service discovery should handle restart due to stale configurations

["Sandor Molnar (Jira)" <ji...@apache.org>]

Sandor Molnar created KNOX-2410:
-----------------------------------

             Summary: CM service discovery should handle restart due to stale configurations
                 Key: KNOX-2410
                 URL: https://issues.apache.org/jira/browse/KNOX-2410
             Project: Apache Knox
          Issue Type: Bug
          Components: cm-discovery
    Affects Versions: 1.4.0
            Reporter: Sandor Molnar
            Assignee: Sandor Molnar
             Fix For: 1.5.0


Cloudera Manager recently changed the event types it triggers when a service is restarted due to stale configurations:
{code:java}
$ curl -ivku admin:admin "https://$CM_HOST:7183/api/v41/events?query=category==AUDIT_EVENT;timeOccurred=gt=2020-07-22T20:10:54.644Z"
...
}, {
    "id" : "5fceaaae-75bb-4be1-92e5-63cb63caf392",
    "content" : "Command RestartWaitingForStalenessSuccess on subject Cluster 1 succeeded.",
    "timeOccurred" : "2020-07-22T20:23:49.784Z",
    "timeReceived" : "2020-07-22T20:23:49.786Z",
    "category" : "AUDIT_EVENT",
    "severity" : "INFORMATIONAL",
    "alert" : false,
    "attributes" : [ {
     ...
    }, {
      "values" : [ "SUCCEEDED" ],
      "name" : "COMMAND_STATUS"
    }, {
      "values" : [ "EV_CLUSTER_RESTARTED" ],
      "name" : "EVENTCODE"
    }, {
      "values" : [ "false" ],
      "name" : "ALERT_SUPPRESSED"
    }, {
      "values" : [ "RestartWaitingForStalenessSuccess" ],
      "name" : "COMMAND"
    }, {
      "values" : [ "ClouderaManager" ],
      "name" : "SERVICE"
    }, {
      "values" : [ "Cluster 1" ],
      "name" : "CLUSTER"
    }, {
      "values" : [ "1" ],
      "name" : "CLUSTER_ID"
    }, {
      "values" : [ "ManagerServer" ],
      "name" : "SERVICE_TYPE"
    }, {
      "values" : [ "COMMAND_SUCCEEDED_WITH_TARGET" ],
      "name" : "MESSAGE_CODES"
    } ] {code}
This new command - {{RestartWaitingForStalenessSuccess}} - should be recognized by Knox and handled the same way as we handle rolling restarts.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)