A suspicion is that the control process seeds a worker process after each job. Something goes awry there, or also possibly, one or the other piece of security software is not liking that process seeding another process, hence why I set exemptions on the BOINC data_dir.

Would be interested to know if this hanging occurs right at a checkpoint save. This can be recorded by adding the <checkpoint_debug>1</checkpoint_debug> flag to the cc_config.xml and setting the *Write to Disk* to default 60 seconds if different. The job slot log info would tell that too, but having it shown in the client message log permanently records this to the stdoutdae.txt file with time-stamp.

--//--

Yes, BOINCTasks shows them, but is not logging them, only telling when the last one was attempted by the science app [which one can see in the task properties as well] and only for as long as BOINCTasks is up and connected to the client. Was hoping that logged data maybe would tell when the freezing started. The gap between Elapsed and CPU time allowing to calculate back to the approximate matching time in the client log.

--//--