-
Notifications
You must be signed in to change notification settings - Fork 25
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[MacOS] Performance observation: Kantra is ~144x slower than WindUp on jboss-eap-quickstarts #121
Comments
This issue is likely a duplicate of #96, I filed it separately though as I thought the magnitude of the difference was worth being explicit and ensuring we didn't lose for tracking. Perhaps this repo (https://github.com/jboss-developer/jboss-eap-quickstarts) is a good test for us to work with to iron out the perf issues. |
@jwmatthews with the change #127 is still taking a long time I didn't let it finish. wonder if the maven packages were already cached on the system, it is taking a long time in download AFAICT Going to see how this works when we are not writing the deps to a maven cache that is a mount to see if it speeds that part of up. |
After removing the .m2 cache and using the change seeing a new time of: I am seeing a 6 min download twice for the deps though so 12 of the min is that. is that acceptable? |
In the future, we may want to allow to use a local .m2 as a volume, but because of the Mac OS and windows issues there I think it would cause more issues than not right now. I wonder if we keep this around as a feature request for that once it does merge? |
@shawn-hurley on my M1 Max with #127 went down to 36 mins (was 192 mins).
|
As of 12/1/2023, on my M1 Max, took ~15 mins to analyze Ticket Monster: My reproducer/output is here: https://github.com/jwmatthews/sample_kantra_reports/tree/main/ticket-monster
|
I ran an analysis on the jboss-eap-quickstarts repo, all of the subdirectories at once (~6k files present in checkout, ~23Mb of content, ~90 small projects ).
Kantra took 192 minutes to complete the analysis.
I re-ran this with WindUp and tried to use similar targets, and it took ~1 minute 20 seconds.
This analysis was run by accident, I intended to only run against kitchensink, but forgot and ran on entire checkout, was then surprised with how long it was taking.
Environment: Mac OS (apple silicon - M1Max)
Reproducer Kantra (using binary from ":latest" image as of Nov 15 2023)
Create podman machine via: https://github.com/jwmatthews/kyma_poc/blob/main/data/darwin_restart_podman_machine.sh
git clone https://github.com/jboss-developer/jboss-eap-quickstarts.git jboss-eap-quickstarts-javaee
time ./kantra analyze -i jboss-eap-quickstarts-javaee -t "quarkus" -t "jakarta-ee" -t "jakarta-ee8+" -t "jakarta-ee9+" -t "cloud-readiness" -o out --overwrite
Reproducer Windup (using MTA 6.2.0 binary)
Reuse checkout of jboss-eap-quickstarts-javaee from above
time ./windup-cli --input jboss-eap-quickstarts-javaee --target jakarta-ee quarkus quarkus3 cloud-readiness --output ./out.windup --overwrite
real 1m21.521s
user 2m39.934s
sys 0m9.134s
159.94s user 9.14s system 207% cpu 1:21.54 total
The text was updated successfully, but these errors were encountered: