Yeah, most of our performance testing showed either no difference (within the
margin of error of Java which can be a handful of percent) or 10%+ improvement.
One thing I might suggest is to double check that the environment is exactly the
same as previous tests, or rerun older tests with your current environment to
make sure it's all the same. We've definitely had cases where our test system
was running out of memory or we accidentally changed the Java version and it
caused a significant performance degradation.
On 2024-10-07 07:44 AM, Mike Beckerle wrote:
Well that's disappointing. Our measurements were more like a 10% improvement
from 3.8.0 to 3.9.0.
Clearly there is more work to do.
On Wed, Oct 2, 2024 at 7:07 AM Roger L Costello <coste...@mitre.org
<mailto:coste...@mitre.org>> wrote:
Time for Daffodil to parse a file containing 5-million records, each record
has 132 characters:
Daffodil version 3.6 took 15 minutes
Daffodil version 3.7 took 12.5 minutes
Daffodil version 3.8 took 12.6 minutes
Daffodil version 3.9 took 13.9 minutes
The performance of Daffodil has steadily declined since version 3.7.
Daffodil is now 11.3% slower than it was in version 3.7.
