[jira] [Commented] (HBASE-9266) Javadoc: Document that HBaseAdmin.flush(...) is synchronous

["Lars Hofhansl (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-9266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13744486#comment-13744486 ] 

Lars Hofhansl commented on HBASE-9266:
--------------------------------------

Note: It is fine to promise an asynchronous API and then actually perform the action synchronously. The reverse is not true.

With that in mind I could a new method to HBaseAdmin: syncFlush(...) which guarantees synchronous operation. That way we can make flush() asynchronous in the future, without breaking any expectation in the caller.

Thoughts?

                
> Javadoc: Document that HBaseAdmin.flush(...) is synchronous
> -----------------------------------------------------------
>
>                 Key: HBASE-9266
>                 URL: https://issues.apache.org/jira/browse/HBASE-9266
>             Project: HBase
>          Issue Type: Bug
>            Reporter: Lars Hofhansl
>            Assignee: Lars Hofhansl
>            Priority: Trivial
>             Fix For: 0.98.0, 0.94.12, 0.95.3
>
>
> Currently the Javadoc in HBaseAdmin.flush(...) says that this is an asynchronous operation. Looking at the, however, it is in fact not.
> Having this as a synchronous operation open ups some use cases we're interested in. For example doing some ingest via the high level API with WAL disabled and then issue a flush through the HBaseAdmin. When the flush succeeded we know the data is securely in HBase.
> (This should work even with intermediary splits, as they also force a flush, but that needs to be verified as well.
> So this is just for a trivial Javadoc change.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira