[
https://issues.apache.org/jira/browse/HBASE-17434?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15810038#comment-15810038
]
Eshcar Hillel commented on HBASE-17434:
---------------------------------------
Making a copy inside the building block methods and extracting them to a
generic solution is similar to using CopyOnWriteXXX instead of LinkedList. This
would duplicate the level of locks in use - one in the CopyOnWriteXXX to
protect the creation of the read-only copy, and one in the compaction pipeline
to ensure atomic access of the linked-list and the version number.
For this reason we implement copy-on-write at this level and we keep the lines
of taking the lock and creating the copy next to each other at the high level
public methods.
> New Synchronization Scheme for Compaction Pipeline
> --------------------------------------------------
>
> Key: HBASE-17434
> URL: https://issues.apache.org/jira/browse/HBASE-17434
> Project: HBase
> Issue Type: Bug
> Reporter: Eshcar Hillel
> Assignee: Eshcar Hillel
> Attachments: HBASE-17434-V01.patch, HBASE-17434-V02.patch
>
>
> A new copyOnWrite synchronization scheme is introduced for the compaction
> pipeline.
> The new scheme is better since it removes the lock from getSegments() which
> is invoked in every get and scan operation, and it reduces the number of
> LinkedList objects that are created at runtime, thus can reduce GC (not by
> much, but still...).
> In addition, it fixes the method getTailSize() in compaction pipeline. This
> method creates a MemstoreSize object which comprises the data size and the
> overhead size of the segment and needs to be atomic.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
