Adapted the tape code.
Moved the Errnum unit tests and covered all the throwers.

to the BaseDaemon class.

opposed to the BaseDaemon class.

to the BaseDaemon class.

BaseDaemon.

the -O2 option.

submission

BaseDaemon.

Pumped up the maximum length of the string value (why do we have one ?) in order to have a better chance to get full backtraces

pointer to Mutex member variable names MultiThreadedDaemon::m_signalMutex.

In case the compilation takes place in debug mode, add code to extract file, function and line numbers in the backtraces

Allow overwriting the build type on the command line via a e.g. '-D CMAKE_BUILD_TYPE:STRING=Debug' option

submitted requests to disappear.
The logic to clean up old requests has been changed: failed requests
are kept respecting the timeout in CastorConfig, successful requests
may be dropped earlier (as in the past) to keep the SubRequest table from
growing too much.
In consequence, the repack reporting has been adapted: the 'migrated'
count is now computed by difference with the other counters, while the
'failed' count can be trusted because of the above.

Partially added signal handling system for the single threaded daemon.
Cleaned up the local daemonize() function.
Added exception throwing helpers for standard error situations.

NotifierThread.

between libcastorvdqmora.so and libcastorvdqmora.so.  These missing
dependencies only become apparent at runtime.

"transparent" modifications to BaseDaemon.  I am trying to fiux the
following segmentation fault:

Program received signal SIGSEGV, Segmentation fault.
0x00002aaaaac8c96a in castor::server::BaseServer::start (this=0x7fffffffe320)
    at /home/murrayc3/castor/checkout/build/RPM/BUILD/castor-2.1.14-5/castor/server/BaseServer.cpp:166
warning: Source file is more recent than executable.
166	    tp->second->init();

actually causes a sgmentation fault for all C++ daemons at run time:

Program received signal SIGSEGV, Segmentation fault.
0x00002aaaaac89d7f in castor::server::BaseDaemon::addNotifierThreadPool (this=0x7fffffffe350, port=5012)
    at /home/murrayc3/castor/checkout/build/RPM/BUILD/castor-2.1.14-5/castor/server/BaseDaemon.cpp:128
warning: Source file is more recent than executable.
128	  m_threadPools['_']->setNbThreads(0)

This commit hopefully fixes the problem.

can be used directly as the super-class of single-threaded daemons.  There is
currently no need for an intermediary SingleThreadedDaemon class.

SLC5 so I unfortunatley have to lean on TeamCity for this.

Conflicts:
	castor/Imakefile

tape bridge and vdqm daemons will eventaully inherit from one of these new
classes.  This commit is a step towards this goal.  Please note that this
commit modifies the castor::rh::Server and castor::server::BaseDaemon classes
in not so desirable ways.  The castor::rh::Server has been modified so that it
does not use the MetricCollector class.  The castor::server::BaseDaemon class
has been modified so that it does not use either the UDPListenerThreadPool or
the MetricCollector class.  I will put back this code in another commit.

Adapted the unit test which did not simulate this behavior.

Removed the detection of Oracle environment for the client only case (Oracle is not required and might not be present).

Moved the building of the unit tests out of the client only compilation as it requires server libraries for linking.

Bumped up the time waited before cheking that the killed child process is gone from 10ms to 100ms as 10ms was too small in some cases (helgrind + slow virtual machine).

user errors.  This commit adds the following macro that helps enforce
that convention and document it in the code.
  "#define LOG_USERERR LOG_NOTICE"