Hide Forgot
Filed from caserun https://tcms.engineering.redhat.com/run/272331/#caserun_11748176 Version-Release number of selected component (if applicable): util-linux-2.23.2-26.el7_2.1 Steps to Reproduce: run the test in Beaker Actual results: :: [ 09:00:19 ] :: Runnning script -c 'logname; who am i' output.txt, with 5 seconds timeout Script started, file is output.txt Script done, file is output.txt logname: no login name Script started, file is output.txt :: [ 09:00:20 ] :: Command ended itself, I am not killing it. /usr/share/beakerlib/testing.sh: line 851: 21757 Terminated eval "sleep $timeout; touch __INTERNAL_TIMEOUT" :: [ BEGIN ] :: Running 'cat output.txt' Script started on Wed 10 Feb 2016 09:00:19 AM EST logname: no login name Script done on Wed 10 Feb 2016 09:00:19 AM EST :: [ PASS ] :: Command 'cat output.txt' (Expected 0, got 0) :: [ BEGIN ] :: Running 'grep 'no login name' output.txt' logname: no login name :: [ FAIL ] :: Command 'grep 'no login name' output.txt' (Expected 1, got 0) Expected results: when running 'script' manually, I get: # script -c "logname; who am i" output.txt Script started, file is output.txt root root pts/2 Feb 10 09:14 Script done, file is output.txt
Seems like regression. What is the last successful version? I don't see any RHEL7 specific change relevant for the command script(1).
it looks like it _never_ passed on RHEL7 so the last version known to work[*] is util-linux-2.13-0.59.el5_8 - the case wasn't used in RHEL6 because there was "util-linux-ng" component instead and it seems we somehow missed this one when adapting our testsuite for that so I will schedule a RHEL6 run now [*] possibly ... I have found that the case didn't treat timeout as an error, so actually I couldn't say it worked, only that it didn't show this exact error
That's strange, it works on command line, but no when executed by Beaker. Would be possible to add strace to the script to see more details?
FTR, as for RHEL6, it seems to work okay with util-linux-ng-2.17.2-12.18.el6 (In reply to Karel Zak from comment #4) > That's strange, it works on command line, but no when executed by Beaker. > Would be possible to add strace to the script to see more details? ok, see job #1223112 (hope it will work out as expected ...)
(In reply to Karel Volný from comment #5) > ok, see job #1223112 (hope it will work out as expected ...) bad luck, second try - job #1224020
Is this still anything we need to fix?
(In reply to Karel Zak from comment #10) > Is this still anything we need to fix? probably I'm not the one to decide what needs to be fixed, however the test fails consistently in Beaker, and I do not understand why it didn't get any attention meanwhile, during previous errata runs RHEL7 results: https://beaker.engineering.redhat.com/jobs/4578810 RHEL8 results: https://beaker.engineering.redhat.com/jobs/4578812 note that there is also strace as agreed before as the test passes manually (or in 1minutetip) and fails only in Beaker, it is not sure whether the problem lies with util-linux, however, from my limited knowledge, I'd think that stdin has to have an owner so it should be printed - if this is not true, I'm eager to hear how to setup the environment so the test really tests what is to be tested but note that the test still passes on RHEL5: https://beaker.engineering.redhat.com/jobs/4581104 so I do not see this as a testbug per se
After evaluating this issue, there are no plans to address it further or fix it in an upcoming release. Therefore, it is being closed. If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened.