[
https://issues.apache.org/jira/browse/GEODE-8349?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17157712#comment-17157712
]
ASF GitHub Bot commented on GEODE-8349:
---------------------------------------
bschuchardt commented on pull request #5363:
URL: https://github.com/apache/geode/pull/5363#issuecomment-658425247
I've written some geode-benchmark tests to vet this PR. The tests add a new
ClusterTopology to the benchmarks so that there are more servers and workload
operations are distributed among the servers. I wrote a replicated region
`get` benchmark but since all data was local it was pretty meaningless and so I
removed it.
```
org.apache.geode.benchmark.tests.PartitionedGetBenchmark
average ops/second Baseline: 415568.87 Test: 421152.13
Difference: +1.3%
ops/second standard error Baseline: 1137.52 Test: 1202.93
Difference: +5.7%
ops/second standard deviation Baseline: 34087.75 Test: 36047.68
Difference: +5.7%
YS 99th percentile latency Baseline: 20007.00 Test: 20007.33
Difference: +0.0%
median latency Baseline: 3619.00 Test: 3539.00
Difference: -2.2%
90th percentile latency Baseline: 822271.00 Test: 823295.00
Difference: +0.1%
99th percentile latency Baseline: 41058303.00 Test: 41746431.00
Difference: +1.7%
99.9th percentile latency Baseline: 96010239.00 Test: 95289343.00
Difference: -0.8%
average latency Baseline: 1732279.46 Test: 1708445.36
Difference: -1.4%
latency standard deviation Baseline: 8040574.12 Test: 8054977.95
Difference: +0.2%
latency standard error Baseline: 416.09 Test: 413.94
Difference: -0.5%
org.apache.geode.benchmark.tests.PartitionedPutBenchmark
average ops/second Baseline: 65766.06 Test: 68103.19
Difference: +3.6%
ops/second standard error Baseline: 280.40 Test: 337.96
Difference: +20.5%
ops/second standard deviation Baseline: 8379.10 Test: 10093.67
Difference: +20.5%
YS 99th percentile latency Baseline: 20080.54 Test: 20080.31
Difference: -0.0%
median latency Baseline: 628735.00 Test: 647679.00
Difference: +3.0%
90th percentile latency Baseline: 31834111.00 Test: 29949951.00
Difference: -5.9%
99th percentile latency Baseline: 173539327.00 Test: 176160767.00
Difference: +1.5%
99.9th percentile latency Baseline: 251789311.00 Test: 271581183.00
Difference: +7.9%
average latency Baseline: 11008644.65 Test: 10651543.96
Difference: -3.2%
latency standard deviation Baseline: 31209391.24 Test: 30946703.39
Difference: -0.8%
latency standard error Baseline: 4070.65 Test: 3970.15
Difference: -2.5%
org.apache.geode.benchmark.tests.ReplicatedPutBenchmark
average ops/second Baseline: 36802.56 Test: 36020.12
Difference: -2.1%
ops/second standard error Baseline: 579.41 Test: 638.69
Difference: +10.2%
ops/second standard deviation Baseline: 16842.76 Test: 18599.04
Difference: +10.4%
YS 99th percentile latency Baseline: 20085.94 Test: 20085.82
Difference: -0.0%
median latency Baseline: 563711.00 Test: 641535.00
Difference: +13.8%
90th percentile latency Baseline: 32653311.00 Test: 29753343.00
Difference: -8.9%
99th percentile latency Baseline: 507772927.00 Test: 596639743.00
Difference: +17.5%
99.9th percentile latency Baseline: 758644735.00 Test: 882376703.00
Difference: +16.3%
average latency Baseline: 20738470.30 Test: 20930653.55
Difference: +0.9%
latency standard deviation Baseline: 82082668.05 Test: 90329383.31
Difference: +10.0%
latency standard error Baseline: 14695.56 Test: 16252.34
Difference: +10.6%
```
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
> reinstate use of SSLSocket for cluster communication
> ----------------------------------------------------
>
> Key: GEODE-8349
> URL: https://issues.apache.org/jira/browse/GEODE-8349
> Project: Geode
> Issue Type: Bug
> Components: membership, messaging
> Reporter: Bruce J Schuchardt
> Assignee: Bruce J Schuchardt
> Priority: Major
>
> We've found problems with "new IO"'s SSLEngine with respect to support for
> TLSV1. We've also seen anomalous performance using that secure
> communications mechanism. The introduction of the use of the "new IO"
> SSLEngine was originally to 1) reduce code complexity in the
> org.apache.geode.internal.tcp package and 2) to set the stage for its use in
> client/server communications so that selectors could be used in c/s
> communications.
> This ticket aims to reintroduce the use of SSLSocket in cluster
> communications without restoring the old, poorly tested SSL code paths. The
> new implementation should have as good or better performance than the
> previous"old IO" implementation and the more recent "new IO" SSLEngine
> implementation as well. This should be apparent in the CI benchmark jobs.
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
