[
https://issues.apache.org/jira/browse/HBASE-17081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15760756#comment-15760756
]
Eshcar Hillel commented on HBASE-17081:
---------------------------------------
bq. enabling BASIC by default for all cases am not sure if that is what is
going to happen when there is really no duplicates in the actual user workload
Basic does not eliminate duplication. It flattens to cell array map and flushes
all segments to disk in one cut.
bq. But what if this test fail indicate some issue?
Obviously if there is a bug in a test then it needs to be fixed.
In the cases that I found the test failed due to assertion failure. For
example, when expecting specific number of items or bytes either in memory or
on disk. Since basic compaction behaves differently the layout in memory or on
disk is different and therefore the assertion fails. In these cases I suggest
to set compaction to none, as this was the initial intention of the test.
bq. But it seems even in any case compacting memstore is in place as default
now.
Since Basic is the default, compacting memstore is the default. But as
explained this could be set by setting one of the properties above.
bq. You refer to which config/setter here? We had a HCD boolean based setter to
enable compacting memstore. Now that is removed against this new Enum type
setter no? Am I missing some thing?
Yes. The boolean flag was replaced with an enum NONE | BASIC | EAGER.
> Flush the entire CompactingMemStore content to disk
> ---------------------------------------------------
>
> Key: HBASE-17081
> URL: https://issues.apache.org/jira/browse/HBASE-17081
> Project: HBase
> Issue Type: Sub-task
> Reporter: Anastasia Braginsky
> Assignee: Anastasia Braginsky
> Attachments: HBASE-15787_8.patch, HBASE-17081-V01.patch,
> HBASE-17081-V02.patch, HBASE-17081-V03.patch, HBASE-17081-V04.patch,
> HBASE-17081-V05.patch, HBASE-17081-V06.patch, HBASE-17081-V06.patch,
> HBASE-17081-V07.patch, HBaseMeetupDecember2016-V02.pptx,
> Pipelinememstore_fortrunk_3.patch
>
>
> Part of CompactingMemStore's memory is held by an active segment, and another
> part is divided between immutable segments in the compacting pipeline. Upon
> flush-to-disk request we want to flush all of it to disk, in contrast to
> flushing only tail of the compacting pipeline.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
