[
https://issues.apache.org/jira/browse/HBASE-20390?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16443470#comment-16443470
]
stack commented on HBASE-20390:
-------------------------------
I tried this. It is a bit better than default on writes but seems worse for
mixed load. For single writes/load (with client-side-buffering enabled), its
27832 vs 26897... about ~3% difference.
I then compared old default to new running load, workloada (50/50) and
workloadc (all read) using the pattern I've been using up to this over in
HBASE-20188 -- except now clients are sending updates in 2MB batches instead of
as a row at a time).
Current defaults ops/second: 25828, 35213, 22000
New defaults ops/second: 25812 (+0%), 27981 (-20%), 20198 (-8%)
I attached a png showing the two runs, first with current defaults and second
with new defaults: hits.ihc.png
!hits.ihc.png!
> IMC Default Parameters for 2.0.0
> --------------------------------
>
> Key: HBASE-20390
> URL: https://issues.apache.org/jira/browse/HBASE-20390
> Project: HBase
> Issue Type: Sub-task
> Reporter: Eshcar Hillel
> Assignee: Eshcar Hillel
> Priority: Major
> Attachments: HBASE-20390-branch-2.0-01.patch,
> HBASE-20390-branch-2.0-01.patch, HBASE-20390.branch-2.0.002.patch,
> hits.ihc.png
>
>
> Setting new default parameters for in-memory compaction based on performance
> tests done in HBASE-20188
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
