[jira] [Created] (AMBARI-13144) Error during update service configurations while kerberizing cluster post Ambari upgrade

["Andrew Onischuk (JIRA)" <ji...@apache.org>]

Andrew Onischuk created AMBARI-13144:
----------------------------------------

             Summary: Error during update service configurations while kerberizing cluster post Ambari upgrade
                 Key: AMBARI-13144
                 URL: https://issues.apache.org/jira/browse/AMBARI-13144
             Project: Ambari
          Issue Type: Bug
            Reporter: Andrew Onischuk
            Assignee: Andrew Onischuk
             Fix For: 2.1.2


Steps followed:  
1\. Install Ambari 1.6.1 with HDP 2.1.15.0-946  
2\. Enable kerberos  
3\. Upgrade Ambari to 2.1.2-262  
4\. Remove Ganglia service via API  
curl -u admin:admin -H 'X-Requested-By:ambari' -X DELETE
'http://172.22.123.214:8080/api/v1/clusters/cl1/services/GANGLIA'  
5\. Add Ambari Metrics Service  
6\. Re–enable Kerberos and choose Existing MIT KDC - specify valid values in
the wizard and navigate till the 'Kerberize Cluster' screen

Result:  
Error at Update Service Configurations phase (see attached screenshot)

ambari-server.log shows below NPE:

    
    
    17 Sep 2015 19:12:23,336 ERROR [Server Action Executor Worker 1325] ClusterImpl:2411 - No service found for config type '{}', service config version not created
    17 Sep 2015 19:12:23,534  WARN [Server Action Executor Worker 1325] ServerActionExecutor:479 - Task #1325 failed to complete execution due to thrown exception: java.lang.NullPointerException:null
    java.lang.NullPointerException
            at java.util.HashMap.putAll(HashMap.java:614)
            at org.apache.ambari.server.state.ConfigHelper.updateConfigType(ConfigHelper.java:691)
            at org.apache.ambari.server.serveraction.kerberos.UpdateKerberosConfigsServerAction.execute(UpdateKerberosConfigsServerAction.java:132)
            at org.apache.ambari.server.serveraction.ServerActionExecutor$Worker.execute(ServerActionExecutor.java:537)
            at org.apache.ambari.server.serveraction.ServerActionExecutor$Worker.run(ServerActionExecutor.java:474)
            at java.lang.Thread.run(Thread.java:745)
    17 Sep 2015 19:12:24,342  WARN [ambari-action-scheduler] ActionScheduler:311 - Operation completely failed, aborting request id:89
    17 Sep 2015 19:12:24,342  INFO [ambari-action-scheduler] ActionScheduler:700 - Service name is , component name is AMBARI_SERVER_ACTIONskipping sending ServiceComponentHostOpFailedEvent for AMBARI_SERVER_ACTION
    17 Sep 2015 19:12:24,346  INFO [ambari-action-scheduler] ActionDBAccessorImpl:176 - Aborting command. Hostname vsharma-u21todalm10-re-5.novalocal role AMBARI_SERVER_ACTION requestId null taskId 1326 stageId null
    17 Sep 2015 19:12:33,092  INFO [qtp-client-22] PersistKeyValueService:82 - Looking for keyName hostPopup-pagination-displayLength-admin
    17 Sep 2015 19:22:33,681  INFO [qtp-client-22] PersistKeyValueService:82 - Looking for keyName hostPopup-pagination-displayLength-admin
    

After discussing with rlevas, turns out that the root cause is because the API
call to delete Ganglia did not properly cleanup the entries database leading
to the issue

ambari=> select serviceconfig.service_name, clusterservices.service_name from
serviceconfig left outer join clusterservices using (service_name) where
clusterservices.service_name is null;  
service_name | service_name  
\-------------<del>+</del>\-------------  
GANGLIA |  
GANGLIA |  
(2 rows)





--
This message was sent by Atlassian JIRA
(v6.3.4#6332)