I am using a docker image from OWASP in my pipeline to scan my web app and produce a HTML report, and I am encountering a problem I've spent the whole day trying to solve.
When running the scan job, it will successfully scan the website, but immediatly after executing the scan command, the job will stop and return "error: job failed: exit code 1".
This happens without precisions on what has failed in the command.
Here is the code of my job :
zap_scan:
stage: owasp
image:
name: owasp/zap2docker-stable
script:
- mkdir /zap/wrk
- /zap/zap-baseline.py -t http://webURL.com -g gen.conf -r /zap/wrk/report.html
artifacts:
paths:
- /zap/wrk/report.html
Normally, this shouldn't be returning an error as a I have tested the scan command by running this on a locally built, identic docker image and I have encountered no issues (ie the scan and the file were generated properly).
Here is the open source code of the zap-baseline.py script
By looking into this, I've found that the script can return error 1 if fail_count is different than 0.
I do not understand why the script behave differently on a local docker image and in a pipeline, can you help me please ?
Gitlab sends an automatic error when a script returns 1, which was the case because the scan had warnings. Adding the option -I fixed the error.