[
https://issues.apache.org/jira/browse/HADOOP-1298?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12551878
]
Doug Cutting commented on HADOOP-1298:
--------------------------------------
I worry about the performance of this patch. Every file action is
synchcronized ,and within those synchronized blocks a new PermissionChecker is
created, the file path is parsed, etc. We must minimize code in synchronized
blocks on the namenode. Ideally the path could be parsed and as many objects
allocated, etc. outside of the sync block.
Must our permission checking be perfectly synchronized? How bad would it be if
groups or permissions were to change within an RPC, between the time that the
permissions are checked and the operation performed? Given the decentralized
nature of Hadoop, I'm not sure we need to aim for perfection here, and it might
be much more efficient to not require perfect synchronization. I'd argue that
the primary goal of synchronization here should be to prevent corrupt or
inconsistent datastructures. If a user is permitted to create a file while
another user is, e.g., removing his permission to do so, that might be okay.
> adding user info to file
> ------------------------
>
> Key: HADOOP-1298
> URL: https://issues.apache.org/jira/browse/HADOOP-1298
> Project: Hadoop
> Issue Type: New Feature
> Components: dfs, fs
> Reporter: Kurtis Heimerl
> Assignee: Christophe Taton
> Attachments: 1298_2007-09-22_1.patch, 1298_2007-10-04_1.patch,
> 1298_20071206b.patch, 1298_20071213.patch, hadoop-user-munncha.patch17
>
>
> I'm working on adding a permissions model to hadoop's DFS. The first step is
> this change, which associates user info with files. Following this I'll
> assoicate permissions info, then block methods based on that user info, then
> authorization of the user info.
> So, right now i've implemented adding user info to files. I'm looking for
> feedback before I clean this up and make it offical.
> I wasn't sure what release, i'm working off trunk.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
