[ 
https://issues.apache.org/jira/browse/HDDS-14237?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Swaminathan Balachandran updated HDDS-14237:
--------------------------------------------
    Description: As per rocksdb DirectSlices are supposed to expose the direct 
buffers to interact with rocksdb. The only problem with directSlices when 
working with external byte buffers it doesn't understand offset and limit of 
the ByteBuffer. Hence to make it safe to work just creating a constructor for 
ManagedDirectSlice fixing the offset and length of the underlying directSlice 
instance should be good. The current implementation doesn't make it safe to 
reuse the direct slice multiple times since rocksdb doesn't allow rollbacks of 
offset in case we want to use a different offset and limit with the same byte 
buffer.  (was: As per rocksdb DirectSlices are supposed to expose the direct 
buffers to interact with rocksdb. The only problem with directSlices when 
working with external byte buffers it doesn't understand offset and limit of 
the ByteBuffer. Hence to make it safe to work just creating a constructor for 
ManagedDirectSlice fixing the offset and length of the underlying directSlice 
instance should be good.)

> Simplify ManagedDirectSlice to use ByteBuffers
> ----------------------------------------------
>
>                 Key: HDDS-14237
>                 URL: https://issues.apache.org/jira/browse/HDDS-14237
>             Project: Apache Ozone
>          Issue Type: Sub-task
>            Reporter: Swaminathan Balachandran
>            Assignee: Swaminathan Balachandran
>            Priority: Major
>
> As per rocksdb DirectSlices are supposed to expose the direct buffers to 
> interact with rocksdb. The only problem with directSlices when working with 
> external byte buffers it doesn't understand offset and limit of the 
> ByteBuffer. Hence to make it safe to work just creating a constructor for 
> ManagedDirectSlice fixing the offset and length of the underlying directSlice 
> instance should be good. The current implementation doesn't make it safe to 
> reuse the direct slice multiple times since rocksdb doesn't allow rollbacks 
> of offset in case we want to use a different offset and limit with the same 
> byte buffer.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to