li-leyang commented on code in PR #4181:
URL: https://github.com/apache/hadoop/pull/4181#discussion_r869785759
##########
hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/viewfs/InodeTree.java:
##########
@@ -682,6 +748,32 @@ protected InodeTree(final Configuration config, final
String viewName,
}
}
+ /**
Review Comment:
Because sorting changes the existing implementation if nested mount point is
disabled(current state) and is also not required with more overhead.
##########
hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/viewfs/InodeTree.java:
##########
@@ -133,6 +138,11 @@ public INode(String pathToNode, UserGroupInformation aUgi)
{
// and is read only.
abstract boolean isInternalDir();
+ /**
+ * INode representing a INodeDir which also contains a INodeLink(nested
mount point)
+ */
+ abstract boolean isDirAndLink();
+
// INode linking to another filesystem. Represented
// via mount table link config entries.
boolean isLink() {
Review Comment:
In particular, I'd say that isLink should return true on all links
(internal & leaf)
> Yeah, I've considered this but this involves refactoring in
traversal/resolve logic and also change the semantics of isLink() which returns
true for an internal node. (I am more concerned of this from traversal
perspective but it might be contradictory from definition perspective) So I
decided to create a new prop to identify internal node with mount point. But
the downside is I have to add additional check whenever x.isLink() is invoked.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
