01-29-2014 09:54 AM
Doing some testing with NBU 7.6.0.1 and running into an issue with OpsCenter. I've got a CentOS 6.5 sysetm that I've installed OpsCenter onto. The installation reported as successful. But, the new ops_atd will not start. It seems to be looking for some libraries in the /opt/VRTSat/plugins directory. Of course, there is no /opt/VRTSat/plugins directory since the "regular" version of VRTSat is not installed any longer.
Started by going into the /opt/SYMCOpsCenterServer/authbroker/bin directory to look at the vxatd.log file. Nothing other than stating "At Init successful". Used the command "./vssat setloglevel -l 4". This generated some additional information:
(7050|140461559543584) At Init successful.
(7050|140461559543584) Initialized local configuration successfully
*** Broker Not In FIPS mode ***
Jan 29 11:21:00 2014:50826,18,0,7050,3244062496,debug,AT,8: (7050|140461559543584) Warning! Unable to Load Unified Logging settings.
Jan 29 11:21:00 2014:50826,18,0,7050,3244062496,debug,AT,8: (7050|140461559543584) Check if Unified Logging configuration exists for ProductID 50826 and OriginatorID 18.
Jan 29 11:21:00 2014:50826,18,0,7050,3244062496,debug,AT,8: (7050|140461559543584) Actual Logging file name doesnt match with bootstrap log file name
Jan 29 11:21:00 2014:50826,18,0,7050,3244062496,debug,AT,8: (7050|140461559543584) Finishing boot strap logging
Jan 29 11:21:00 2014:50826,18,0,7050,3244062496,debug,AT,8: (7050|140461559543584) Detailed broker log would be available at /var/VRTSat/log/vxatd.log
Note the reference to /var/VRTSat/log/vxatd.log. I created that file, ran './vssat setbrokerlog -l 4" and attempted to start the ops_atd again (by running /opt/SYMCOpsCenterServer/bin/startat". This created the following information in /var/VRTSat/log/vxatd.log:
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: ACE_DLL_Manager_Ex::open: Invalid handle: /opt/VRTSat/plugins/libauthldap.so: cannot open shared object file: No such file or directory
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: ACE_DLL_Manager::open_dll: Could not open dll.
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: (7262|140306184017696) Unable to load dll /opt/VRTSat/plugins/libauthldap.so in pluggins.cpp(100)
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,3: (7262|140306184017696) DoNotLoad=true for library /opt/VRTSat/plugins/libauthsecureid.so. Plugin library wont be loaded.
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: ACE_DLL_Manager_Ex::open: Invalid handle: /opt/VRTSat/plugins/libauthlocalhost_t.so: cannot open shared object file: No such file or directory
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: ACE_DLL_Manager::open_dll: Could not open dll.
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: (7262|140306184017696) Unable to load dll /opt/VRTSat/plugins/libauthlocalhost_t.so in pluggins.cpp(100)
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: (7262|140306184017696) Error loading pluggins in server.cpp(792) program will exit
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: (7262|140306184017696) Broker Raised Exception!
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: (7262|140306184017696) Error: 24631 - Unable to load pluggins in server.cpp(793)
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,3: (7262|140306184017696) retVal: 24631
Jan 29 11:32:03 2014:50826,18,0,7262,2487359264,debug,AT,8: (7262|140306184017696) Broker shutdown complete
Note all the invalid handle messages trying to start plugins in /opt/VRTSat.
Curious if anyone is seeing similar issues on a RedHat system?
I know that CentOS is not officially supported for OpsCenter but if it looks/smells/runs like RedHat then it should be close enough to do some playing with.
Solved! Go to Solution.
02-13-2014 12:40 PM
I had the same issue on RedHat 6.5. I installed ksh (yum install ksh) and restarted the server and all came up. Apparently, some of the authentication setup scripts require ksh, which is why the ops_atd daemon wasn't setting up properly.
PAF
02-13-2014 03:16 AM
Same issue here.
Fresh installation of OpsCenter 7.6.0.1 with all defaults, cannot log in. Tried both Red Hat and SuSE - same results.
02-13-2014 12:40 PM
I had the same issue on RedHat 6.5. I installed ksh (yum install ksh) and restarted the server and all came up. Apparently, some of the authentication setup scripts require ksh, which is why the ops_atd daemon wasn't setting up properly.
PAF
02-13-2014 01:32 PM
That helped, thanks!
02-13-2014 02:19 PM
Logicalis_PAF, great catch. I had to re-install OpsCenter on my CentOS 6.5 server after installing ksh. That did the trick.
02-13-2014 04:51 PM
Thats good news. Unfortunately, mine was an upgrade of 7.5.0.5 OCA, so I needed to get it working. Frustratingly enough, the server services were running fine, we just couldn't log in to see all the wonderful new features in 7.6.0.1. I'm glad that I could share a fix for the community that has helped me so much in the past.
Cheers.
PAF
02-19-2014 08:43 AM
thanks!
a nother dependency is glibc.i686 by the way.