narendly commented on code in PR #2072:
URL: https://github.com/apache/helix/pull/2072#discussion_r860842513
##########
helix-core/src/main/java/org/apache/helix/HelixProperty.java:
##########
@@ -39,7 +39,7 @@
* A wrapper class for ZNRecord. Used as a base class for IdealState,
CurrentState, etc.
*/
public class HelixProperty {
Review Comment:
I took a cursory look and decided to add a thought here based on the context
I have, so correct me if I'm wrong or missed any other details:
Because we "deep copy" `HelixProperty` and its internal `ZNRecord`,
throughout Helix we can treat this as such, and that is, we can assume that
it's immutable. But in the case that we do not clone it and simply take after
the internal `ZNRecord` reference, such assumptions no longer hold and
therefore may open up avenues where issues could creep in. And this particular
change doesn't allow us to know whether its internal `ZNRecord` reference is
being shared across different `HelixProperty` instances or not.
Could we think of alternative ways to achieve this if you think this is
necessary? For example, add a subclass of `HelixProperty` that shallow copies
`ZNRecord`, so that there is no ambiguity?
--
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]
