[
https://issues.apache.org/jira/browse/EAGLE-997?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jayesh updated EAGLE-997:
-------------------------
Description:
currently eagle docker image is based on ambari, which while deploying requires
entire hadoop cluster to be built by ambari shell API as a part of building
docker image.
if we use docker image that is not only based on prebuilt hadoop but also has
required eagle dependencies, the only thing that needs be done is to deploy
eagle by following the deployment steps.
this gives lot of benefits
1) deployment steps can be verified when there is a change,
2) it serves as a indirect documentation on working deployment steps
3) makes it easy to test any change on eagle code ( before CI )
4) provides effective dev environment so use does have to spend time to install
all required components.
was:
currently eagle docker image is based on ambari, which while deploying requires
entire hadoop cluster to be built by ambari shell API as a part of building
docker image.
if we use docker image that is based on prebuilt hadoop, the only thing that
needs be done is to deploy eagle by following the deployment steps.
this gives lot of benefits
1) deployment steps can be verified when there is a change,
2) it serves as a indirect documentation on working deployment steps
3) makes it easy to test any change on eagle code ( before CI )
4) provides effective dev environment so use does have to spend time to install
all required components.
> use prebuilt hadoop base image for eagle docker image
> -----------------------------------------------------
>
> Key: EAGLE-997
> URL: https://issues.apache.org/jira/browse/EAGLE-997
> Project: Eagle
> Issue Type: Improvement
> Affects Versions: v0.5.0
> Reporter: Jayesh
> Assignee: Jayesh
> Fix For: v0.6.0
>
>
> currently eagle docker image is based on ambari, which while deploying
> requires entire hadoop cluster to be built by ambari shell API as a part of
> building docker image.
> if we use docker image that is not only based on prebuilt hadoop but also has
> required eagle dependencies, the only thing that needs be done is to deploy
> eagle by following the deployment steps.
> this gives lot of benefits
> 1) deployment steps can be verified when there is a change,
> 2) it serves as a indirect documentation on working deployment steps
> 3) makes it easy to test any change on eagle code ( before CI )
> 4) provides effective dev environment so use does have to spend time to
> install all required components.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
