[
https://issues.apache.org/jira/browse/AMBARI-20560?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Christopher Jackson updated AMBARI-20560:
-----------------------------------------
Description:
It appears that HDP 2.6 will support both Spark1 and Spark2. It will allow
these services to be installed side by side and an environment variable will
determine which is used when a process that needs spark is run.
Services have a metainfo.xml file which they can list other services which are
required for their service to function properly. If a service required Spark
(any version) how would they define this in the metainfo.xml file? For instance
I wouldn’t want the following as it would imply that both Spark 1.x and Spark
2.x had to be installed:
{code}
<requiredServices>
<service>SPARK</service>
<service>SPARK2</service>
</requiredServices>
{code}
I also wouldn’t want to list only one or the other because this service
supports both versions.
There should be some sort of mechanism to
was:
It appears that HDP 2.6 will support both Spark1 and Spark2. It will allow
these services
to be installed side by side and an environment variable will determine which
is used when a process
that needs spark is run.
Services have a metainfo.xml file which they can list other services which are
required for
their service to function properly. If a service required Spark (any version)
how would they
define this in the metainfo.xml file? For instance I wouldn’t want the
following as it would
imply that both Spark 1.x and Spark 2.x had to be installed:
{code}
<requiredServices>
<service>SPARK</service>
<service>SPARK2</service>
</requiredServices>
{code}
I also wouldn’t want to list only one or the other because this service
supports both versions.
There should be some sort of mechanism to
> Need mechanism to add one of a set of services as required in metainfo.xml
> file.
> ---------------------------------------------------------------------------------
>
> Key: AMBARI-20560
> URL: https://issues.apache.org/jira/browse/AMBARI-20560
> Project: Ambari
> Issue Type: Story
> Reporter: Christopher Jackson
>
> It appears that HDP 2.6 will support both Spark1 and Spark2. It will allow
> these services to be installed side by side and an environment variable will
> determine which is used when a process that needs spark is run.
> Services have a metainfo.xml file which they can list other services which
> are required for their service to function properly. If a service required
> Spark (any version) how would they define this in the metainfo.xml file? For
> instance I wouldn’t want the following as it would imply that both Spark 1.x
> and Spark 2.x had to be installed:
> {code}
> <requiredServices>
> <service>SPARK</service>
> <service>SPARK2</service>
> </requiredServices>
> {code}
> I also wouldn’t want to list only one or the other because this service
> supports both versions.
> There should be some sort of mechanism to
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
