[
https://issues.apache.org/jira/browse/HBASE-5230?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13191606#comment-13191606
]
Phabricator commented on HBASE-5230:
------------------------------------
mbautin has commented on the revision "[jira] [HBASE-5230] Extend
TestCacheOnWrite to ensure we don't cache data blocks on compaction".
Responses to comments inline.
INLINE COMMENTS
src/main/java/org/apache/hadoop/hbase/io/hfile/LruBlockCache.java:746 Done.
src/main/java/org/apache/hadoop/hbase/regionserver/Store.java:756 Done.
src/main/java/org/apache/hadoop/hbase/regionserver/metrics/SchemaMetrics.java:880
Replaced this with a method that "prints out" the metrics into a StringBuilder
and returns a string.
src/test/java/org/apache/hadoop/hbase/io/hfile/TestCacheOnWrite.java:253
Since this is for data blocks, I renamed this to
testNotCachingDataBlocksDuringCompaction.
src/main/java/org/apache/hadoop/hbase/regionserver/Store.java:764-766 That is
correct, to my best knowledge. Eventually, I think we would like to
intelligently decide whether to cache-on-write a block based on whether the
data in question is already in the block cache as part of uncompacted files:
https://issues.apache.org/jira/browse/HBASE-5263
src/main/java/org/apache/hadoop/hbase/regionserver/metrics/SchemaMetrics.java:879-880
In my debugging I have not yet come across a case when I would have found a
SchemaMetrics.toString method useful. Also, adapting this static method to
implement toString would be tricky, since it relies on getMetricsSnapshot()
that takes a "snapshot" of _all_ metrics, not just those for a particular
table/CF combination corresponding to one SchemaMetrics instance. Therefore, I
would prefer to leave SchemaMetrics.toString() out for now.
src/test/java/org/apache/hadoop/hbase/client/TestFromClientSide.java:4502
Done.
src/test/java/org/apache/hadoop/hbase/io/hfile/TestCacheOnWrite.java:254
Added.
src/test/java/org/apache/hadoop/hbase/io/hfile/TestCacheOnWrite.java:295 Done.
REVISION DETAIL
https://reviews.facebook.net/D1353
> Unit test to ensure compactions don't cache data on write
> ---------------------------------------------------------
>
> Key: HBASE-5230
> URL: https://issues.apache.org/jira/browse/HBASE-5230
> Project: HBase
> Issue Type: Test
> Reporter: Mikhail Bautin
> Assignee: Mikhail Bautin
> Priority: Minor
> Attachments: D1353.1.patch, D1353.2.patch, D1353.3.patch,
> D1353.4.patch,
> Don-t-cache-data-blocks-on-compaction-2012-01-21_00_53_54.patch,
> Don-t-cache-data-blocks-on-compaction-2012-01-23_10_23_45.patch
>
>
> Create a unit test for HBASE-3976 (making sure we don't cache data blocks on
> write during compactions even if cache-on-write is enabled generally
> enabled). This is because we have very different implementations of
> HBASE-3976 without HBASE-4422 CacheConfig (on top of 89-fb, created by Liyin)
> and with CacheConfig (presumably it's there but not sure if it even works,
> since the patch in HBASE-3976 may not have been committed). We need to create
> a unit test to verify that we don't cache data blocks on write during
> compactions, and resolve HBASE-3976 so that this new unit test does not fail.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
