EDIT: Sorry, I accidentally clicked submit before finishing this post.
I reported a similar issue during the 6.1 beta (see topic here: http://forums.thinkboxsoftware.com/viewtopic.php?f=156&t=10463).
The same issue seems to be occurring with the deadlinecommand processes spawned by Pulse. Basically, I had Pulse running for a few days, and then shut it down via deadlinepulse -shutdown from another terminal. Pulse was not able to cleanly close, and after timing out, it eventually resorted to killing itself. This may or may not be related, but this kind of thing is frequent when trying to cleanly close Pulse:
2014-03-18 11:16:08: Stopping Pulse
2014-03-18 11:16:38: Pulse failed to exit after 30 seconds, killing it...
I also saw one or two instances of this in the logs:
2014-03-17 17:44:56: Web Service - Web Service shutting down...
2014-03-17 17:45:55: Info Thread: Stopped
2014-03-17 17:45:55: Clean Up Thread: Stopped
2014-03-17 17:45:55: Pending Job Scan Thread: ShuttingDown
2014-03-17 17:45:55: Server Thread: Stopped
2014-03-17 17:45:55: Scheduler Thread: Stopped
2014-03-17 17:45:55: Power Management: Stopped
2014-03-17 17:45:55: Forcing shutdown.
2014-03-17 17:45:55: Exception Details
2014-03-17 17:45:55: Exception -- One or more threads failed to quit successfully.
2014-03-17 17:45:55: Exception.Data: ( )
2014-03-17 17:45:55: Exception.StackTrace:
2014-03-17 17:45:55: (null)
2014-03-17 17:45:55: Could not query child process information for pid 9702 because: Thread was being aborted (System.Threading.ThreadAbortException)
2014-03-17 17:45:55: WARNING: an error occured while trying to kill the process tree: Thread was being aborted (System.Threading.ThreadAbortException)
2014-03-17 17:45:55: Could not query child process information for pid 9702 because: Thread was being aborted (System.Threading.ThreadAbortException)
2014-03-17 17:45:55: WARNING: an error occured while trying to kill the process tree: Thread was being aborted (System.Threading.ThreadAbortException)
2014-03-17 17:45:55: Error running pending job scan process: Thread was being aborted (System.Threading.ThreadAbortException)
Anyway, after Pulse finished offing itself, I checked the Mongo server to make sure everything had cleaned itself up properly, only to find there were still 4 open connections to the DB. In my process tree, I found 3 orphaned deadlinecommand processes that were previously parented under the deadlinepulse process, and were hanging indefinitely. Two of them were executing deadlinecommand -DoHouseCleaning 10 False, and one was executing deadlinecommand -DoPendingJobScan False. All of them had been running for around 2 days, and after terminating them, all of the remining Mongo connections were properly closed.
This is in 6.2 beta 3 on Fedora 19.