Thanks for the fix Andreas. The SRU bug described well the issue and how to reproduce/test it. I am not sure what was your rationale to rule out the creation of syslog user in postinst which would be the other possible option, however, I agree with the proposed solution, being more cautious and checking if the user exists before using it seems reasonable to me.
I did test it in a kinetic container and the package in your PPA does fix the issue (following the Test Plan section of the bug). I am not testing all the others since the change is pretty much the same.
Thanks for the fix Andreas. The SRU bug described well the issue and how to reproduce/test it. I am not sure what was your rationale to rule out the creation of syslog user in postinst which would be the other possible option, however, I agree with the proposed solution, being more cautious and checking if the user exists before using it seems reasonable to me.
I did test it in a kinetic container and the package in your PPA does fix the issue (following the Test Plan section of the bug). I am not testing all the others since the change is pretty much the same.
FWIW DEP-8 tests are also passing:
autopkgtest [16:23:45]: @@@@@@@ @@@@@@@ @@@@@@ summary
zebra-lo PASS
bgpd-snmp-rpki PASS
py-frr-reload PASS