Thanks a lot for the report. I definitely know what the problem is, and was a bit scared of this :) You can use something such -Dsurefire.stackTraceMaxFrames=-1 I would say that every higher value is useful for JUnit 5+ when parallel execution is enabled. Long story short, the goal is to fix this https://github.com/apache/maven-surefire/issues/3215 But in your case, using Junit4, it's not necessary. I will try to find a way to improve this.
On Sat, 6 Jun 2026 at 16:51, Stanimir Stamenkov via users <[email protected]> wrote: > > Sat, 6 Jun 2026 14:37:40 +1000, /Olivier Lamy/: > > On Sat, 6 Jun 2026 at 03:52, Stanimir Stamenkov via users > > <[email protected]> wrote: > > > >> Comparing 3.5.6 and 3.6.0-M1, I'm noticing a very apparent slow down > >> with (JUnit 4) tests that produce larger output (500K-1MB, > >> maven.test.redirectTestOutputToFile=true). > > > > Does the slowdown happen with maven.test.redirectTestOutputToFile=false? > > Can you create some issue with such report? Anything to reproduce > > would be great. > > redirectTestOutputToFile=false appears to add a fixed overhead with both > versions: > > maven-surefire-plugin | redirectTestOutputToFile | time > ----------------------|--------------------------|------: > 3.5.6 | true | 17s > 3.5.6 | false | 37s > 3.6.0-M1 | true | 57s > 3.6.0-M1 | false | 1m 17s > > So doesn't appear a factor. I've opened: > > - https://github.com/apache/maven-surefire/issues/3373 > > with the example I'm trying with. > > -- > Stanimir > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
