[
https://issues.apache.org/jira/browse/GEODE-2055?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jinmei Liao updated GEODE-2055:
-------------------------------
Description:
1. First, the GeodePermissionResolver [23] is necessary to configure Apache
Shiro's provided (OOTB) Realms correctly. Otherwise, the security Permissions
are not enforced properly (in a hierarchical fashion as advertised [24], i.e.
in section "3. Introduction of ResourcePermission").
I used [25] the GeodePermissionResolver class to configure the Apache Shiro
provided (OOTB) PropertiesRealm implementation [18].
Therefore, the GeodePermissionResolver class must NOT be internal. This is
particularly important if the user is using Apache Shiro to the fullest extent
to configure and secure Apache Geode.
Of course, I could have provided my own implementation of the Apache Shiro
PermissionResolver interface [26] (especially given the simplicity of the
GeodePermissionResolver implementation) but if that implementation every
involves more logic behind the scenes, better to "reuse" then "reinvent" in
this case.
> Expose GeodePermissionResolver
> ------------------------------
>
> Key: GEODE-2055
> URL: https://issues.apache.org/jira/browse/GEODE-2055
> Project: Geode
> Issue Type: Sub-task
> Reporter: Jinmei Liao
>
> 1. First, the GeodePermissionResolver [23] is necessary to configure Apache
> Shiro's provided (OOTB) Realms correctly. Otherwise, the security
> Permissions are not enforced properly (in a hierarchical fashion as
> advertised [24], i.e. in section "3. Introduction of ResourcePermission").
> I used [25] the GeodePermissionResolver class to configure the Apache Shiro
> provided (OOTB) PropertiesRealm implementation [18].
> Therefore, the GeodePermissionResolver class must NOT be internal. This is
> particularly important if the user is using Apache Shiro to the fullest
> extent to configure and secure Apache Geode.
> Of course, I could have provided my own implementation of the Apache Shiro
> PermissionResolver interface [26] (especially given the simplicity of the
> GeodePermissionResolver implementation) but if that implementation every
> involves more logic behind the scenes, better to "reuse" then "reinvent" in
> this case.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
