[
https://issues.apache.org/jira/browse/SENTRY-1103?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15255535#comment-15255535
]
Colin Ma commented on SENTRY-1103:
----------------------------------
[~singhashish], yes, it will be included in how to integrate component with
Sentry. For Kafa, you can update the KafkaPrivilegeModel and test cases for
case sensitive. For example:
cluster with case insensitive:
{code}
implyMethodMap.put(KafkaAuthorizable.AuthorizableType.CLUSTER.name().toLowerCase(),
ImplyMethodType.STRING);
{code}
cluster with case sensitive:
{code}
implyMethodMap.put(KafkaAuthorizable.AuthorizableType.CLUSTER.name().toLowerCase(),
ImplyMethodType.STRING_CASE_SENSITIVE);
{code}
> Authorizable names' case sensitivity must be decided by plugins
> ---------------------------------------------------------------
>
> Key: SENTRY-1103
> URL: https://issues.apache.org/jira/browse/SENTRY-1103
> Project: Sentry
> Issue Type: Sub-task
> Affects Versions: 1.8.0
> Reporter: Ashish K Singh
> Assignee: Colin Ma
> Labels: integration
> Fix For: 1.8.0
>
> Attachments: SENTRY-1103.001-SENTRY-999.patch,
> SENTRY-1103.002-SENTRY-999.patch
>
>
> With more and more components relying on sentry for authorization, it really
> makes sense to have the case-sensitivity configurable. We can keep the
> current behavior as default. However, for use cases like Kafka, it makes
> sense to have authorizable names case sensitive. For instance, a topic
> resource in Kafka can have case sensitive values, like, topic1 and TOPIC1 are
> two completely different valid topic resource instances.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
