[ 
https://issues.apache.org/jira/browse/HBASE-17434?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15815057#comment-15815057
 ] 

Eshcar Hillel commented on HBASE-17434:
---------------------------------------

With the possibility of non-atomic initialization of a LinkedList I added the 
volatile + an explanation of why it is needed for each of the variables.
I wonder -- could it be that there are other variables in the code that have 
the same problem? might be worth checking ...

With respect to the interleaving scenario - I would not call these traps. 
CompctionPipeline is a concurrent data structure that was design for a specific 
purpose; the user should understand its guarantees. One way to show that the 
implementation is correct (linearizable) is to find a point in time during the 
execution of the operation in which it can be linearized. In the above scenario 
there is a point in time during the execution of the isEmpty() where the 
pipeline is empty so the result is valid.

> 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, 
> HBASE-17434-V03.patch, HBASE-17434-V04.patch, HBASE-17434-V05.patch, 
> HBASE-17434.master.001.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)

Reply via email to