elek commented on a change in pull request #2357:
URL: https://github.com/apache/ozone/pull/2357#discussion_r670503213
##########
File path: hadoop-ozone/interface-client/src/main/proto/OmClientProtocol.proto
##########
@@ -528,6 +529,12 @@ enum StorageTypeProto {
RAM_DISK = 4;
}
+enum BucketTypeProto {
Review comment:
Thanks for the suggestion and for the patch @aryangupta1998 . I already
shared my concern offline, but let me repeat m y arguments here in public:
While I really like the concept of having different type of buckets (or more
preciously: different type of behavior for different buckets), I am not sure if
the best representation is introducing a new bucket "type".
We already have a custom `Map<String,String>` field, which (today) is used
to store the flags about FSO or non-FSO behavior.
Having both the custom fields (aka. bucket properties/behaviors) and the
bucket type seems to be confusing for me from usability point of view.
I have no problem with using bucket type on CLI but wouldn't be enough to
implement it only on the client side and use the good old `Map<String,String>`
properties?
It would be more backward and forward compatible (using enum makes it harder
to introduce new values without compatibility issues.)
What do you think?
--
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]
