Comment 30 for bug 1892358

ok, very sorry for my delay in looking at this.

a quick scan of the results for x/b/f show:

ddstreet@smaug:~$ autopkgtest-manager --distro systemd -a all -r xenial -cv --since 2month --minimum 6
parsed '2month' as '2020-08-09T14:53:00.893145-04:00'
88: boot-smoke FAIL non-zero exit status 1
    (amd64:87 arm64:1)
39: systemd-fsckd FAIL non-zero exit status 1
    (amd64:39)
47: boot-and-services FAIL non-zero exit status 1
    (amd64:41 arm64:6)
ddstreet@smaug:~$ autopkgtest-manager --distro systemd -a all -r bionic -cv --since 2month --minimum 6
parsed '2month' as '2020-08-09T14:53:18.194358-04:00'
132: systemd-fsckd FAIL non-zero exit status 1
     (i386:14 amd64:117 ppc64el:1)
15: upstream FAIL non-zero exit status 1
    (i386:3 amd64:10 ppc64el:1 arm64:1)
29: boot-smoke FAIL non-zero exit status 1
    (i386:2 amd64:27)
ddstreet@smaug:~$ autopkgtest-manager --distro systemd -a all -r focal -cv --since 2month --minimum 6
parsed '2month' as '2020-08-09T14:56:04.065710-04:00'
96: systemd-fsckd FAIL non-zero exit status 137
    (amd64:68 ppc64el:20 s390x:8)
21: systemd-fsckd FAIL non-zero exit status 1
    (amd64:1 ppc64el:4 s390x:16)

I think the systemd-fsckd failures are issues with how the test is trying to detect 'good' and 'bad' service status, but in any case i rewrote some of the details of how the test works to be simpler and (hopefully) more accurate/reliable, I'm running it through tests from my ppa now, before I upload the test change.

I'll take a look at the other test failures; the boot-and-services test has typically had a lot of trouble with services randomly not starting ok, but i'll check on the details.