Red Hat Bugzilla – Bug 987056
Condor fails to start
Last modified: 2013-07-22 11:54:03 EDT
Description of problem:
When starting condor:
# systemctl enable condor.service
# systemctl start condor.service
The service does not start. The following error message appears in /var/log/messages every 5 seconds:
Jul 22 16:32:25 hopf systemd: Starting Condor Distributed High-Throughput-Computing...
Jul 22 16:32:25 hopf systemd: Started Condor Distributed High-Throughput-Computing.
Jul 22 16:32:25 hopf systemd: condor.service: main process exited, code=exited, status=4/NOPERMISSION
Jul 22 16:32:25 hopf systemd: condor.service: control process exited, code=exited status=1
Jul 22 16:32:25 hopf systemd: Unit condor.service entered failed state.
Jul 22 16:32:25 hopf condor_master: ERROR "Programmer error: get_mySubSystem() info is invalid(MASTER,0,INVALID)!" at line 1673 in file /builddir/build/BUILD/condor-7.9.5/src/condor_daemon_core.V6/daemon_core_main.cpp
Jul 22 16:32:25 hopf condor_master: SubsystemInfo: name=MASTER type=INVALID(0) class=None(0)
Jul 22 16:32:25 hopf condor_off: Can't find address for local master
Jul 22 16:32:25 hopf condor_off: Perhaps you need to query another pool.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. systemctl start condor.service
See error message listed above. Condor service not running
No error messages in /var/log/messages. Condor service running.
When installing condor :
# rpm -Uvh condor-7.9.5-0.3.fc19.x86_64.rpm condor-classads-7.9.5-0.3.fc19.x86_64.rpm condor-procd-7.9.5-0.3.fc19.x86_64.rpm
The output is as follows:
restorecon: lstat(/var/lock/condor) failed: No such file or directory
libsemanage.dbase_llist_set: record not found in the database (No such file or directory).
libsemanage.dbase_llist_set: could not set record value (No such file or directory).
Could not change boolean condor_domain_can_network_connect
Could not change policy booleans
/sbin/semanage: Port tcp/12345 already defined
The configuration file in /etc/condor/config.d is the distribution one.
/etc/hosts contains the full name for the host
There is already a build to address this issue, and tickets around SELinux which I hope to get to today.
*** This bug has been marked as a duplicate of bug 984061 ***