[
https://issues.apache.org/jira/browse/HBASE-4450?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13109695#comment-13109695
]
[email protected] commented on HBASE-4450:
------------------------------------------------------
bq. On 2011-09-21 16:31:40, Jonathan Gray wrote:
bq. > Sweet! Great test and regression catcher. Not sure if you wanted to
commit this or just put it up to show. Might actually be good to commit so
then when the changes go in the test has just the diffs of behavior changes so
it's easier to understand what it changes/optimizes?
jgray: Yes, I would like to commit it separately, and then update with each of
the fixes (like for hbase-4433, and 4434).
bq. On 2011-09-21 16:31:40, Jonathan Gray wrote:
bq. >
https://svn.apache.org/repos/asf/hbase/trunk/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksRead.java,
line 1
bq. > <https://reviews.apache.org/r/1998/diff/1/?file=44582#file44582line1>
bq. >
bq. > license
will do.
bq. On 2011-09-21 16:31:40, Jonathan Gray wrote:
bq. >
https://svn.apache.org/repos/asf/hbase/trunk/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksRead.java,
line 43
bq. > <https://reviews.apache.org/r/1998/diff/1/?file=44582#file44582line43>
bq. >
bq. > update dir name
will do.
bq. On 2011-09-21 16:31:40, Jonathan Gray wrote:
bq. >
https://svn.apache.org/repos/asf/hbase/trunk/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksRead.java,
lines 157-158
bq. > <https://reviews.apache.org/r/1998/diff/1/?file=44582#file44582line157>
bq. >
bq. > method takes cf but then doesn't use it?
In future, once the per CF/per block type metric is ported to open source
branch, I would like to change the implementation to lookup
cf.<cfName>.Data.fsBlockReadCnt. So would like to leave the cf argument for now.
- Kannan
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/1998/#review1999
-----------------------------------------------------------
On 2011-09-21 09:10:27, Kannan Muthukkaruppan wrote:
bq.
bq. -----------------------------------------------------------
bq. This is an automatically generated e-mail. To reply, visit:
bq. https://reviews.apache.org/r/1998/
bq. -----------------------------------------------------------
bq.
bq. (Updated 2011-09-21 09:10:27)
bq.
bq.
bq. Review request for Michael Stack and Jonathan Gray.
bq.
bq.
bq. Summary
bq. -------
bq.
bq. Test deliberately creates a table with blocksize = 1Byte (so that each KV
is a block of its own).
bq. The getData() calls all take a last argument which is the number of
expected blocks read. Those numbers are the current baseline. I'll hope to
submit patches for hbase-4433 & hbase-4434 in the next day or so, and at that
time these counts should go down.
bq.
bq.
bq. This addresses bug HBASE-4450.
bq. https://issues.apache.org/jira/browse/HBASE-4450
bq.
bq.
bq. Diffs
bq. -----
bq.
bq.
https://svn.apache.org/repos/asf/hbase/trunk/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksRead.java
PRE-CREATION
bq.
bq. Diff: https://reviews.apache.org/r/1998/diff
bq.
bq.
bq. Testing
bq. -------
bq.
bq.
bq. Thanks,
bq.
bq. Kannan
bq.
bq.
> test for number of blocks read: to serve as baseline for expected blocks read
> and for catching regressions
> ----------------------------------------------------------------------------------------------------------
>
> Key: HBASE-4450
> URL: https://issues.apache.org/jira/browse/HBASE-4450
> Project: HBase
> Issue Type: Test
> Reporter: Kannan Muthukkaruppan
> Assignee: Kannan Muthukkaruppan
>
> Add a simple test for number of blocks read. The tests intent is to serve as
> baseline for expected blocks read and for catching regressions. As
> optimizations for HBase-4433 or Hbase-4434 are committed, the test would need
> to be updated to adjust the counts for expected blocks read in various cases.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
