[
https://issues.apache.org/jira/browse/GEODE-6087?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dan Smith updated GEODE-6087:
-----------------------------
Summary: SPIKE: Explore the best way to store and visualize performance
data across many builds (was: Explore the best way to store and visualize
performance data across many builds)
> SPIKE: Explore the best way to store and visualize performance data across
> many builds
> --------------------------------------------------------------------------------------
>
> Key: GEODE-6087
> URL: https://issues.apache.org/jira/browse/GEODE-6087
> Project: Geode
> Issue Type: New Feature
> Components: benchmarks
> Reporter: Dan Smith
> Priority: Major
>
> We want to be able to notice performance regressions in geode and track down
> when they show up. We will want to add running the benchmarks into CI,
> storing the results, and visualizing them.
> We should explore what the best way is to store and visualize the data from
> many benchmark runs.
> A few storage possibilities:
> * Store data as directories or zip files, maybe upload them somewhere?
> * Store data in a relational database?
> * Some combination of the two (derby/sqlite database per run, perhaps?)
> Visualization options:
> * A chart of relative throughput or latency vs. build number?
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
