Description of problem:
The test /CoreOS/tcsh/Sanity/prompt works fine in RHTS but hangs
(for no obvious reason) in Beaker. It uses python pty module for
terminal simulation which could be connected to the issue, but I
Similar for /CoreOS/vim/Sanity/upstream-tests.
Possible the same issue is for /CoreOS/authconfig/sanity/options,
which hangs after the first test phase.
Fine in http://rhts.redhat.com/cgi-bin/rhts/jobs.cgi?id=160588
Suggesting this to be a blocker.
These 2 tests can be only a tip of the iceberg, may affect other tests with the same base.
Similar hang in /CoreOS/perl-DBD-Pg/Sanity/upstream-tests at the
end of the setup phase. Fine in RHTS.
Another RHTS-Beaker difference is for /CoreOS/procps/Sanity/top:
While in RHTS testing "top -m -n 1" works fine, in Beaker we get:
TERM environment variable not set.
:: [ PASS ] :: Checking for the -m option (BZ#479703)
:: [ FAIL ] :: File 'output' should contain 'PID.*USER.*PR.*NI.*USED.*RES.*SHR'
Sounds really terminal related...
/CoreOS/tcsh/Sanity/upstream difference: Four failures appear in
upsteam test suite, all pass in RHTS.
71: onintr FAILED (commands.at:936)
87: source FAILED (commands.at:1206)
112: Comments FAILED (lexical.at:36)
155: $ edit FAILED (variables.at:322)
Will look into that...
This is not Beaker/RHEL6 specific, the same problem appears when
the /CoreOS/perl-DBD-Pg/Sanity/upstream-tests test is run on RHEL5:
Scheduled test run: https://beaker.engineering.redhat.com/jobs/5194
Which failed before with "failed tty get" in TJ#4167.
Please, disregard my last comment, it concerns bug 607994.
Verified to be fixed in 0.5.45. The following tests no more hang
Thanks Peter for verifying it.