[
https://issues.apache.org/jira/browse/ACCUMULO-210?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Eric Newton updated ACCUMULO-210:
---------------------------------
Description:
If you use the .rpm or .deb install packages, the base package is installed.
If you try to run accumulo, you find it is not fully configured, and the start
fails. There's no "slaves" file or "masters".... worse, the helpful start
script tries to create single-node versions of these files. The package isn't
very useful until this is done.
We should layer packages to enable a node to run a given feature:
tablet/logger, master, monitor, tracer. Installing the package will pull in
the base package as a dependency and turn on the feature by installing a file
somewhere. The start-here script should then read those features.
I'm sure hadoop or hbase are doing something similar: at least investigate good
practices.
ese additional steps, basically writing some scripts that work from some
configuration files in /etc, that would be nice.
The goal is to do something like:
{noformat}
$ apt-get install accumulo-single-instance
{noformat}
Or
{noformat}
$ apt-get install accumulo-tserver accumulo-logger
{noformat}
was:
If you use the .rpm or .deb install packages, the base package is installed.
If you try to run accumulo, you find it is not fully configured, and the start
fails. There's no "slaves" file or "masters".... worse, the helpful start
script tries to create single-node versions of these files. The package isn't
very useful until this is done.
We should layer packages to enable a node to run a given feature:
tablet/logger, master, monitor, tracer. Installing the package will pull in
the base package as a dependency and turn on the feature by installing a file
somewhere. The start-here script should then read those features.
I'm sure hadoop or hbase are doing something similar: at least investigate good
practices.
> create packages that enable accumulo features on the given node
> ---------------------------------------------------------------
>
> Key: ACCUMULO-210
> URL: https://issues.apache.org/jira/browse/ACCUMULO-210
> Project: Accumulo
> Issue Type: Bug
> Components: start
> Reporter: Eric Newton
> Assignee: John Vines
> Priority: Blocker
> Fix For: 1.5.0
>
>
> If you use the .rpm or .deb install packages, the base package is installed.
> If you try to run accumulo, you find it is not fully configured, and the
> start fails. There's no "slaves" file or "masters".... worse, the helpful
> start script tries to create single-node versions of these files. The
> package isn't very useful until this is done.
> We should layer packages to enable a node to run a given feature:
> tablet/logger, master, monitor, tracer. Installing the package will pull in
> the base package as a dependency and turn on the feature by installing a file
> somewhere. The start-here script should then read those features.
> I'm sure hadoop or hbase are doing something similar: at least investigate
> good practices.
> ese additional steps, basically writing some scripts that work from some
> configuration files in /etc, that would be nice.
> The goal is to do something like:
> {noformat}
> $ apt-get install accumulo-single-instance
> {noformat}
> Or
> {noformat}
> $ apt-get install accumulo-tserver accumulo-logger
> {noformat}
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
