-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/7115/#review11592
-----------------------------------------------------------

Ship it!


I committed this yesterday.

- Hari Shreedharan


On Sept. 14, 2012, 7:28 p.m., Brock Noland wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/7115/
> -----------------------------------------------------------
> 
> (Updated Sept. 14, 2012, 7:28 p.m.)
> 
> 
> Review request for Flume.
> 
> 
> Description
> -------
> 
> This patch addresses the issue by moving the logic of whether or not to use 
> fast full replay up to the Log class. The reason this is required is because 
> even for a non-existing checkpoint upon start, the file itself will be be 
> created when we create the BackingStore for FlumeEvent queue, which must be 
> done before we create the CheckpointBuilder.
> 
> 
> This addresses bug FLUME-1583.
>     https://issues.apache.org/jira/browse/FLUME-1583
> 
> 
> Diffs
> -----
> 
>   
> flume-ng-channels/flume-file-channel/src/main/java/org/apache/flume/channel/file/CheckpointRebuilder.java
>  6e1d2fc 
>   
> flume-ng-channels/flume-file-channel/src/main/java/org/apache/flume/channel/file/Log.java
>  1d91460 
>   
> flume-ng-channels/flume-file-channel/src/main/java/org/apache/flume/channel/file/ReplayHandler.java
>  7c32526 
>   
> flume-ng-channels/flume-file-channel/src/test/java/org/apache/flume/channel/file/TestCheckpointRebuilder.java
>  PRE-CREATION 
>   
> flume-ng-channels/flume-file-channel/src/test/java/org/apache/flume/channel/file/TestFileChannelRestart.java
>  4133573 
>   
> flume-ng-channels/flume-file-channel/src/test/java/org/apache/flume/channel/file/TestUtils.java
>  2b88b96 
> 
> Diff: https://reviews.apache.org/r/7115/diff/
> 
> 
> Testing
> -------
> 
> There is no automated way to test whether a fast full replay occurred. I have 
> added two tests
> 
> testFastReplayNegativeTestV1
> testFastReplayNegativeTestV2
> 
> which can be used to manually verify the logic. When run you look for
> 
> 
> Found NextFileID 0, from []
> Not disabling fast full replay because checkpoint  does not exists: 
> .../chkpt/checkpoint
> 
> Found NextFileID 3, from [....]
> Disabling fast full replay because checkpoint exists: .../chkpt/checkpoint
> 
> in the debug output. I think we can remove the need for manual testing via 
> FLUME-1580.
> 
> 
> Thanks,
> 
> Brock Noland
> 
>

Reply via email to