You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ambari.apache.org by "Martin Bohacek (JIRA)" <ji...@apache.org> on 2017/03/03 12:40:45 UTC

[jira] [Updated] (AMBARI-20298) Delete Service in Ambari 2.4 does not work properly

     [ https://issues.apache.org/jira/browse/AMBARI-20298?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Martin Bohacek updated AMBARI-20298:
------------------------------------
    Attachment: add_service_errors.log

> Delete Service in Ambari 2.4 does not work properly
> ---------------------------------------------------
>
>                 Key: AMBARI-20298
>                 URL: https://issues.apache.org/jira/browse/AMBARI-20298
>             Project: Ambari
>          Issue Type: Bug
>          Components: ambari-sever
>    Affects Versions: 2.4.2
>            Reporter: Martin Bohacek
>         Attachments: add_service_errors.log
>
>
> Deleting a service using Ambari UI (Service Actions > Delete Service) does not delete the service completely.
> We have custom service for Fluentd, on which I observe the problem. After using deletion from UI, whe infor about the service still remains in Ambari database tables - in our case e.g.:
> {noformat}
> ambari=> select * from clusterservices where service_name like 'FLUENT%';
>  service_name | cluster_id | service_enabled 
> --------------+------------+-----------------
>  FLUENTD      |          2 |               0
> (1 row)
> ambari=> select * from servicecomponentdesiredstate where component_name like 'FLUENT%';
>  id  |          component_name          | cluster_id | desired_stack_id | desired_version | desired_state | service_name | recovery_enabled 
> -----+----------------------------------+------------+------------------+-----------------+---------------+--------------+------------------
>  107 | FLUENTD_KNOX_GATEWAY_COLLECTOR   |          2 |                3 | UNKNOWN         | INSTALLED     | FLUENTD      |                0
>  101 | FLUENTD_CONSUMER                 |          2 |                3 | UNKNOWN         | STARTED       | FLUENTD      |                0
>  102 | FLUENTD_HDFS_NAMENODE_COLLECTOR  |          2 |                3 | UNKNOWN         | INSTALLED     | FLUENTD      |                0
>  103 | FLUENTD_HIVE_SERVER2_COLLECTOR   |          2 |                3 | UNKNOWN         | INSTALLED     | FLUENTD      |                0
>  104 | FLUENTD_SYSTEM_COLLECTOR         |          2 |                3 | UNKNOWN         | INSTALLED     | FLUENTD      |                0
>  105 | FLUENTD_FORWARDER                |          2 |                3 | UNKNOWN         | INSTALLED     | FLUENTD      |                0
>  106 | FLUENTD_HIVE_METASTORE_COLLECTOR |          2 |                3 | UNKNOWN         | INSTALLED     | FLUENTD      |                0
> {noformat}
> This is problem, when I want to redeploy the service again - error occurs on database integrity check level (see attached stack trace from Ambari server log).
> To fix the problem, there is a workaround to manually clean the database tables after performing service deletion from web UI:
> {noformat}
> delete from hostcomponentdesiredstate where component_name like 'FLUENT%';
> delete from hostcomponentstate where component_name like 'FLUENT%';
> delete from servicecomponentdesiredstate where component_name like 'FLUENT%';
> delete from servicedesiredstate where service_name like 'FLUENT%';
> delete from clusterservices where service_name like 'FLUENT%';
> {noformat}



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)