[jira] [Resolved] (RANGER-2052) The Ranger admin supports high availability

["Pradeep Agrawal (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/RANGER-2052?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Pradeep Agrawal resolved RANGER-2052.
-------------------------------------
    Fix Version/s: 2.1.0
       Resolution: Implemented

Ranger admin HA is already implemented via RANGER-2555. 

MySQL HA is completely not possible from Ranger end as it has to sync the diff after failed node comes live. Better MySQL HA solution should be try out. 

However if you still want to address then please create new Jira and will track the implementation there.

> The Ranger admin supports high availability
> -------------------------------------------
>
>                 Key: RANGER-2052
>                 URL: https://issues.apache.org/jira/browse/RANGER-2052
>             Project: Ranger
>          Issue Type: New Feature
>          Components: admin
>            Reporter: Qiang Zhang
>            Assignee: Qiang Zhang
>            Priority: Major
>              Labels: newbie, patch
>             Fix For: 2.1.0
>
>
> In real business environment, such as banks application and so on,  we should use other Ranger service to continuely supply service once  a Ranger service exits abnormally. We call the feature as high availability. The Ranger high availability includes:
> 1. The Ranger admin is high availability;
> 2. The mysql service that saves meta data is high availability.
> For the Ranger admin high availability our scheme is as following:
> 1. Add a configuration item in plugin. The configuration item includes two the address of the Ranger admin service.
> 2. The plugin refreshed policy from other Ranger admin service once it refresh policy from a Ranger admin service fail.
> For the mysql service that saves meta data admin high availability, our scheme is as following:
> 1. Add a configuration item in the Ranger admin. The configuration item includes two the address of the mysql service.
> 2. Build two jdbc connection pools. The other will be used once a jdbc connection fail.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)