[
https://issues.apache.org/jira/browse/HBASE-9775?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Elliott Clark updated HBASE-9775:
---------------------------------
Attachment: Charts Search Cloudera Manager.png
Here's the network. The client barely reaches 1.2 gbit of a possible 10 gbit.
We're also seeing a pretty huge slow down vs 0.94. Something in the order of
15% slower on the write path (the write path is better or the same).
So I'm thinking that it's got to be something in the AsyncProcess code.
Batching seems suspect but I'm running more tests to check. For me that huge a
pref loss and a scaling loss are pretty huge issues.
> Client write path scales very badly with more servers
> -----------------------------------------------------
>
> Key: HBASE-9775
> URL: https://issues.apache.org/jira/browse/HBASE-9775
> Project: HBase
> Issue Type: Bug
> Components: Client
> Affects Versions: 0.96.0
> Reporter: Elliott Clark
> Attachments: Charts Search Cloudera Manager.png, short_ycsb.png
>
>
> Testing on larger clusters has not had the desired throughput increases.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
