[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Attachment: patched.workloadw.80percentwrites.json
notpatched.workloadw.80percentwrites.json
[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Attachment: compare.png
bq. We might be able to get that 5% back by going with the complexity of
rollbacks on
[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Attachment: ops.png
gc.png
Here is perf run with the latest patch from HBASE-15082: i.e. wait
[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Attachment: 50threads_ycsb.png
I tried with 50 ycsb threads running through same suite. The graphs look samish
[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Attachment: call_times_0-1_and_1-3_ycsb.png
Here is run of a suite of ycsb tests (30 threads to a single
[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Attachment: writes.png
latencies.png
all_under_lock.svg
[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Assignee: stack
> Perf test doing all mutation steps under row lock
>
[
https://issues.apache.org/jira/browse/HBASE-15046?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-15046:
--
Attachment: 1.2.svg
1.2.v2.svg
Here is what a RS looks like when YCSB write only is being run
