[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrew Purtell updated HBASE-15392:
---
Fix Version/s: (was: 1.4.0)
> Single Cell Get reads two HFileBlocks
>
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Resolution: Fixed
Fix Version/s: 1.4.0
1.3.0
Status: Resolved (was: Patch
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Status: Patch Available (was: Reopened)
> Single Cell Get reads two HFileBlocks
>
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392.backport.branch-1.patch
Backport
> Single Cell Get reads two HFileBlocks
>
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Resolution: Fixed
Hadoop Flags: Reviewed
Fix Version/s: 2.0.0
Status: Resolved (was:
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v9.patch
v9 is v7 w/ just the two lines changed.. moving '>=' to '>'.
This is what I pushed
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: HBASE-15265-v8.patch
Address [~lhofhansl] last comment
> Single Cell Get reads two HFileBlocks
>
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v7 (1).patch
Retry. See if timeouts have gone away
> Single Cell Get reads two HFileBlocks
>
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v7.patch
The jdk8 tests failed because of outstanding TestSimpleRpcScheduler failure.
This
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Release Note:
When an explicit Get with a one or more columns specified, we at a minimum,
were overseeking,
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v7.patch
Retry. I unchecked RUN_IN_DOCKER
> Single Cell Get reads two HFileBlocks
>
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v7.patch
Some hadoopqa builds are running... others are getting this docker fail
retry
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v7.patch
Retry
> Single Cell Get reads two HFileBlocks
>
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v7.patch
Address [~lhofhansl] comment. Retry. See if docker succeeds this time.
> Single
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v6.patch
-v6 includes your comment w/ an edit of what was in optimize. Will commit if
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v5.patch
v5. Doc not done yet because I am not yet sure on the index compare. Also need
to
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: reads.png
gc.png
io.png
Small YCSB test comparing before and after
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: reads.png
gc.png
Two charts comparing GC and hits w/ patch and w/o. Weird is how
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v4.patch
Patch with tests. Optimize seems broke to me. We keep SKIPping even though
we've
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ramkrishna.s.vasudevan updated HBASE-15392:
---
Attachment: HBASE-15392_suggest.patch
Just attaching a suggestion patch.
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Lars Hofhansl updated HBASE-15392:
--
Attachment: 15392-0.98-looksee.txt
Hey [~stack], does -looksee fix the problem for you (sorry
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v3.wip.patch
So, here is v3. Still does not have tests (though it sounds like we need lots
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: no_optimize.patch
no_optimize.patch
Ok. If I comment out the optimize method, I
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: two_seeks.txt
Here is log when stuff is broke -- we are getting two blocks. The query is:
{code}
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Release Note: When an explicit Get with a one or more columns specified, we
at a minimum, were overseeking
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Release Note: When an explicit Get with a one or more columns specified, we
at a minimum, were overseeking
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392v2.wip.patch
Retry. The failed test was actually a wrong path. Removed. Made the patch
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Status: Patch Available (was: Open)
See what we break with this patch.
> Single Cell Get reads two
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Attachment: 15392.wip.patch
WIP
ScanQueryMatcher takes a Scan and copies locally a bunch of its state -- i.e.
[
https://issues.apache.org/jira/browse/HBASE-15392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15392:
--
Description:
As found by Daniel "SystemTap" Pol, a simple Get results in our reading two
HFileBlocks, the one
30 matches
Mail list logo
