You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ambari.apache.org by "Tim Thorpe (JIRA)" <ji...@apache.org> on 2016/02/19 17:21:18 UTC
[jira] [Commented] (AMBARI-14494) Generate warning on deploy wizard
UI when HAWQ segment is not co-located with a datanode.
[ https://issues.apache.org/jira/browse/AMBARI-14494?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15154446#comment-15154446 ]
Tim Thorpe commented on AMBARI-14494:
-------------------------------------
There seems to be a problem with this fix for Spark. SPARK_THRIFTSERVER and HIVE_SERVER are component names NOT service names. So they will never appear in the servicesList.
+ servicesList = [service["StackServices"]["service_name"] for service in services["services"]]
+ if "SPARK" in servicesList:
+ if "SPARK_THRIFTSERVER" in servicesList:
+ if not "HIVE_SERVER" in servicesList:
+ servicesList = [service["StackServices"]["service_name"] for service in services["services"]]
+ componentsList = [item["StackServiceComponents"] for sublist in componentsListList for item in sublist]
The following line should be added after the componentsList calculation:
componentNames = [component["component_name"] for component in componentsList]
and the checks for SPARK_THRIFTSERVER and HIVE_SERVER should be changed to:
+ if "SPARK_THRIFTSERVER" in componentNames:
+ if not "HIVE_SERVER" in componentNames:
> Generate warning on deploy wizard UI when HAWQ segment is not co-located with a datanode.
> -----------------------------------------------------------------------------------------
>
> Key: AMBARI-14494
> URL: https://issues.apache.org/jira/browse/AMBARI-14494
> Project: Ambari
> Issue Type: Improvement
> Components: ambari-server
> Affects Versions: trunk, 2.2.0
> Reporter: Lav Jain
> Assignee: Lav Jain
> Priority: Minor
> Fix For: 2.3.0, 2.2.1
>
> Attachments: AMBARI-14494-2.branch22.patch, AMBARI-14494-2.patch, AMBARI-14494.patch, Assignments.png, Warning.png
>
>
> HAWQ can run on nodes that without hdfs deployed, that is access data remotely. However, there are significant performance improvements if segments and datanodes are co-located.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)