[
https://issues.apache.org/jira/browse/HDDS-9?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16690979#comment-16690979
]
Xiaoyu Yao commented on HDDS-9:
-------------------------------
Thanks [~jnp] for the review. The block ID is a string in block token
identifier.
After HDDS-749, BCSID is already part of BlockID class, when instantiate
identifier from BlockToken Manager, we can use the BlockID#toString() to
include CID:LID:BCSID. The current container implementation already have logic
to reject bcsid greater than container's bcsid without block token.
We can also add length as suggested. My concern is that it will restrict the
block token usage to stale data even new writes have been committed. The client
will need to get new token even it is not expired to read the newly written
data, this may add additional load on the OM BlockToken Manager.
> Add GRPC protocol interceptors for Ozone Block Token
> ----------------------------------------------------
>
> Key: HDDS-9
> URL: https://issues.apache.org/jira/browse/HDDS-9
> Project: Hadoop Distributed Data Store
> Issue Type: Sub-task
> Components: Security
> Reporter: Xiaoyu Yao
> Assignee: Xiaoyu Yao
> Priority: Major
> Attachments: HDDS-9-HDDS-4.001.patch, HDDS-9-HDDS-4.002.patch
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
