[
https://issues.apache.org/jira/browse/METRON-386?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15430710#comment-15430710
]
ASF GitHub Bot commented on METRON-386:
---------------------------------------
GitHub user justinleet opened a pull request:
https://github.com/apache/incubator-metron/pull/223
METRON-386 Create Ambari Service Definition for Elasticsearch
https://issues.apache.org/jira/browse/METRON-386
Opening the PR to get community feedback.
Prereq:
Ambari 2.4
Testing this PR:
*Copy or mount the services and stack contents into the appropriate places
on the ambari-server node. For example:
**
incubator-metron/metron-deployment/packaging/ambari/src/main/resources/stacks/HDP/2.4/services/ELASTICSEARCH
-> /var/lib/ambari-server/resources/stacks/HDP/2.4/services/ELASTICSEARCH
**incubator-metron/metron-deployment/packaging/ambari/src/main/resources/common-services/ELASTICSEARCH
-> /var/lib/ambari-server/resources/common-services/ELASTICSEARCH
* ES Master eligible nodes will not contain data. Don't colocate master
eligible nodes and data nodes.
* It's necessary to provide the ES zen_discovery_ping_unicast_hosts.
Limitations / Notes
* This is based off of Symantec's work at:
https://github.com/Symantec/ambari-elasticsearch-service. Should I add this
into the NOTICE that includes similar contributions?
* Most configs are used from Symantec's build. If there are more sane
configs, we should change the defaults (I'm not an ES expert by any means).
* It would be really nice if zen_discovery_ping_unicast_hosts could be
automatically populated with something sane. I'm unsure of how to accomplish
this with Ambari.
* Master-eligible nodes are non-data holding nodes that are eligible to be
elected master. Data nodes are, not surprisingly, nodes that only hold data,
but are not master-eligible. It would be nice if thing could be modified to
allow master eligible nodes to also hold data cleanly, but this is left as a
further enhancement if needed.
* A service advisor definition could probably be added to force Ambari to
avoid placing master-eligible and data nodes onto the same node.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/justinleet/incubator-metron
metron-elasticsearch
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-metron/pull/223.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #223
----
commit 9861d95e45e5c9b3b39fea0722baafd3bb5d0dc1
Author: justinjleet <[email protected]>
Date: 2016-08-18T13:04:11Z
Elasticsearch Service in Ambari
----
> Create Ambari Service Definition for Elasticsearch
> --------------------------------------------------
>
> Key: METRON-386
> URL: https://issues.apache.org/jira/browse/METRON-386
> Project: Metron
> Issue Type: New Feature
> Reporter: Justin Leet
> Assignee: Justin Leet
>
> Spinning up Elasticsearch through Ambari would make deployment and management
> a lot easier and comprehensive.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
