[
https://issues.apache.org/jira/browse/DAFFODIL-1876?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dave Thompson closed DAFFODIL-1876.
-----------------------------------
Pulled latest updates from incubator-daffodil repository and verified the
specified commit, ee4177b4c9bf84af2af19d8c9f67f283e233105e, was included.
Reviewed updated files to determine the effected tests. Executed the entire SBT
test suite. Verified that the effected tests executed successfully. Nightly
test run executed successfully.
Comparison with the FormatInfo commit (a1d881), showed performance gains across
all data format tests with some exceptions. Some of these could be due to
normal run-to-run deviation. There were no nato-stanag tests listed in this
comparison due to these test being broken at the time.
The comparison to the pre- FormatInfo commit (401761) showed that performance
has not been regained to pre-FormatInfo commit values across all format tests.
Most significant are the army-drss and nato-stanag tests which needs further
investigation.
Also note that for some tests there is a significant performance increase from
the 2.0.0 baseline and in some there is still a significant performance
decrease.
Closing this ticket and created JIRA ticket DAFFODIL-1883 to identify the
specific data formats that need further performance regression investigation.
> Improve performance regression due to FormatInfo changes
> --------------------------------------------------------
>
> Key: DAFFODIL-1876
> URL: https://issues.apache.org/jira/browse/DAFFODIL-1876
> Project: Daffodil
> Issue Type: Improvement
> Components: Back End, Middle "End"
> Affects Versions: 2.1.0
> Reporter: Michael Beckerle
> Assignee: Dave Thompson
> Priority: Major
> Fix For: 2.1.0
>
>
> FormatInfo changes introduced significant performance degradation.
> Changes are needed to win back this performance.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
