[
https://issues.apache.org/jira/browse/HBASE-13082?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15181578#comment-15181578
]
ramkrishna.s.vasudevan commented on HBASE-13082:
------------------------------------------------
[~lhofhansl]
In the current trunk and 1.3 actually we are not doing any updateReaders after
compactions. Only after flushes we are doing. So how does this patch work with
flushes if you will wait on the latch that is created per StoreScanner. Because
for flush we are forced to update the scanner because we are cleaning up the
memstore snapshot. But yes we do have the reference to the older memstore
snapshot but holding it up for a longer time is not good is what was discussed
previously in this JIRA impl. So even if compactions get blocked if there are
large scans then it will affect the compaction requests that comes newly?
> Coarsen StoreScanner locks to RegionScanner
> -------------------------------------------
>
> Key: HBASE-13082
> URL: https://issues.apache.org/jira/browse/HBASE-13082
> Project: HBase
> Issue Type: Bug
> Components: Performance, Scanners
> Reporter: Lars Hofhansl
> Assignee: ramkrishna.s.vasudevan
> Fix For: 2.0.0
>
> Attachments: 13082-test.txt, 13082-v2.txt, 13082-v3.txt,
> 13082-v4.txt, 13082.txt, 13082.txt, CountDownLatch-0.98.txt, HBASE-13082.pdf,
> HBASE-13082_1.pdf, HBASE-13082_12.patch, HBASE-13082_13.patch,
> HBASE-13082_14.patch, HBASE-13082_15.patch, HBASE-13082_16.patch,
> HBASE-13082_17.patch, HBASE-13082_18.patch, HBASE-13082_19.patch,
> HBASE-13082_1_WIP.patch, HBASE-13082_2.pdf, HBASE-13082_2_WIP.patch,
> HBASE-13082_3.patch, HBASE-13082_4.patch, HBASE-13082_9.patch,
> HBASE-13082_9.patch, HBASE-13082_withoutpatch.jpg, HBASE-13082_withpatch.jpg,
> LockVsSynchronized.java, gc.png, gc.png, gc.png, hits.png, next.png, next.png
>
>
> Continuing where HBASE-10015 left of.
> We can avoid locking (and memory fencing) inside StoreScanner by deferring to
> the lock already held by the RegionScanner.
> In tests this shows quite a scan improvement and reduced CPU (the fences make
> the cores wait for memory fetches).
> There are some drawbacks too:
> * All calls to RegionScanner need to be remain synchronized
> * Implementors of coprocessors need to be diligent in following the locking
> contract. For example Phoenix does not lock RegionScanner.nextRaw() and
> required in the documentation (not picking on Phoenix, this one is my fault
> as I told them it's OK)
> * possible starving of flushes and compaction with heavy read load.
> RegionScanner operations would keep getting the locks and the
> flushes/compactions would not be able finalize the set of files.
> I'll have a patch soon.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
