[OE-core] [PATCH] logrotate.py: restore /etc/logrotate.d/wtmp

mingli.yu at windriver.com mingli.yu at windriver.com
Wed Feb 13 05:43:28 UTC 2019

From: Mingli Yu <Mingli.Yu at windriver.com>

During the test logrotate.LogrotateTest.test_1_logrotate_setup,
there is below logic:
 # mkdir $HOME/logrotate_dir
 # sed -i "s#wtmp {#wtmp {\n    olddir $HOME/logrotate_dir#" /etc/logrotate.d/wtmp

After all logrotate.LogrotateTest finished, only cleanup
$HOME/logrotate_dir as below, but don't restore
the config file /etc/logrotate.d/wtmp.
def tearDownClass(cls):
    cls.tc.target.run('rm -rf $HOME/logrotate_dir')

That's to say, there is one additional line added
to /etc/logrotate.d/wtmp and will make the logrotate
service start failed when run systemd.SystemdBasicTests.test_systemd_failed

Take an example as below when run test as root:
 # cat /etc/logrotate.d/wtmp
 # no packages own wtmp -- we'll rotate it here
 /var/log/wtmp {
    olddir /root/logrotate_dir
    create 0664 root utmp
    minsize 1M
    rotate 1

 # ls /root/logrotate_dir
 ls: cannot access '/root/logrotate_dir': No such file or directory

 # systemctl start logrotate
 Job for logrotate.service failed because the control process exited with error code.
 See "systemctl status logrotate.service" and "journalctl -xe" for details.

 # systemctl status logrotate
  logrotate.service - Rotate log files
   Loaded: loaded (/lib/systemd/system/logrotate.service; static; vendor preset>
   Active: failed (Result: exit-code) since Wed 2019-02-13 03:35:19 UTC; 7s ago
     Docs: man:logrotate(8)
   Process: 540 ExecStart=/usr/sbin/logrotate /etc/logrotate.conf (code=exited, status=1/FAILURE)
   Main PID: 540 (code=exited, status=1/FAILURE)

  Feb 13 03:35:18 qemumips systemd[1]: Starting Rotate log files...
  Feb 13 03:35:19 qemumips logrotate[540]: error: wtmp:9 error verifying olddir path /root/logrotate_dir: No such file or directory
  Feb 13 03:35:19 qemumips logrotate[540]: error: found error in file wtmp, skipping
  Feb 13 03:35:19 qemumips systemd[1]: logrotate.service: Main process exited, code=exited, status=1/FAILURE
  Feb 13 03:35:19 qemumips systemd[1]: logrotate.service: Failed with result 'exit-code'.
  Feb 13 03:35:19 qemumips systemd[1]: Failed to start Rotate log files.

Add the logic to restore /etc/logrotate.d/wtmp to
make the cleanup complete to fix the above issue.

Signed-off-by: Mingli Yu <Mingli.Yu at windriver.com>
 meta/lib/oeqa/runtime/cases/logrotate.py | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/meta/lib/oeqa/runtime/cases/logrotate.py b/meta/lib/oeqa/runtime/cases/logrotate.py
index db6e695..d266644 100644
--- a/meta/lib/oeqa/runtime/cases/logrotate.py
+++ b/meta/lib/oeqa/runtime/cases/logrotate.py
@@ -9,8 +9,12 @@ from oeqa.runtime.decorator.package import OEHasPackage
 class LogrotateTest(OERuntimeTestCase):
+    def setUpClass(cls):
+        cls.tc.target.run('cp /etc/logrotate.d/wtmp $HOME/wtmp.oeqabak')
+    @classmethod
     def tearDownClass(cls):
-        cls.tc.target.run('rm -rf $HOME/logrotate_dir')
+        cls.tc.target.run('mv -f $HOME/wtmp.oeqabak /etc/logrotate.d/wtmp && rm -rf $HOME/logrotate_dir')

More information about the Openembedded-core mailing list