[
https://issues.apache.org/jira/browse/HBASE-4148?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13073236#comment-13073236
]
[email protected] commented on HBASE-4148:
------------------------------------------------------
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/1229/#review1243
-----------------------------------------------------------
Ship it!
KVs come in descending timestamp order so that the most recent data is always
the first thing on disk when you seek to a particular row-col. This is because
the default "get" mode wants to see the latest version, rather than an older
one. So, you have to skip over less data.
As for tests passing, so long as this new test passes, don't worry about the
others. You can check the Apache hudson instance to see what has failed
recently.
src/test/java/org/apache/hadoop/hbase/mapreduce/TestHFileOutputFormat.java
<https://reviews.apache.org/r/1229/#comment2840>
assertEquals
- Todd
On 2011-07-30 18:40:49, jmhsieh wrote:
bq.
bq. -----------------------------------------------------------
bq. This is an automatically generated e-mail. To reply, visit:
bq. https://reviews.apache.org/r/1229/
bq. -----------------------------------------------------------
bq.
bq. (Updated 2011-07-30 18:40:49)
bq.
bq.
bq. Review request for hbase and Todd Lipcon.
bq.
bq.
bq. Summary
bq. -------
bq.
bq. When HFiles are flushed through the normal path, they include an attribute
TIMERANGE_KEY which can be used to cull HFiles when performing a
time-restricted scan. Files produced by HFileOutputFormat are currently missing
this metadata.
bq.
bq.
bq. This addresses bug HBASE-4148.
bq. https://issues.apache.org/jira/browse/HBASE-4148
bq.
bq.
bq. Diffs
bq. -----
bq.
bq. src/main/java/org/apache/hadoop/hbase/mapreduce/HFileOutputFormat.java
8ccdf4d
bq. src/main/java/org/apache/hadoop/hbase/regionserver/StoreFile.java
40efdda
bq.
src/test/java/org/apache/hadoop/hbase/mapreduce/TestHFileOutputFormat.java
89241eb
bq.
bq. Diff: https://reviews.apache.org/r/1229/diff
bq.
bq.
bq. Testing
bq. -------
bq.
bq. Added unit test.
bq.
bq. I don't quite understand why the KeyValue with the larger timestamp (2000)
value must be written before the one with the smaller timestamp (1000). I can
see the code that enforces this (HFile.checkKey) but not why keys are larger to
smaller. Is this in HFile data precondition?
bq.
bq. I cannot get the full test suite to pass, with or without this patch.
Suite seems to timeout on tests unrelated to this. Would appreciate some hints
or pointers for info on which tests are flakey or take a long time to run.
bq.
bq.
bq. Thanks,
bq.
bq. jmhsieh
bq.
bq.
> HFileOutputFormat doesn't fill in TIMERANGE_KEY metadata
> --------------------------------------------------------
>
> Key: HBASE-4148
> URL: https://issues.apache.org/jira/browse/HBASE-4148
> Project: HBase
> Issue Type: Bug
> Components: mapreduce
> Affects Versions: 0.90.3
> Reporter: Todd Lipcon
> Assignee: Jonathan Hsieh
> Fix For: 0.90.5
>
> Attachments:
> 0001-HBASE-4148.-HFileOutputFormat-doesn-t-fill-in-TIMERA.patch
>
>
> When HFiles are flushed through the normal path, they include an attribute
> TIMERANGE_KEY which can be used to cull HFiles when performing a
> time-restricted scan. Files produced by HFileOutputFormat are currently
> missing this metadata.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
