[sheepdog] killing sheep is unsafe - sheep: add option to create pid file

[Jens WEBER]

related problems
   - Bug: pidfile option? [main] crash_handler(408) sheep pid 31144 exited unexpectedly.
   - Bug: strange behavior after restart sheeps - potential loss of data

sheep doesn't respond correctly on kill TERM (15) signal and crash. I think this is the reason of my problem and potential cause lost of data. 

It would be nice that sheep terminate on signal 15 like on shutdown. (like Dietmar mentioned before)

really bad shutdown
root at sheep01:~# /etc/init.d/sheepdog stop
[ ok ] Stopping sheepdog: sheepdog_1.
Jul 18 12:21:48 [main] crash_handler(408) sheep pid 21337 exited unexpectedly.

nice shutdown
root at sheep01:~# collie cluster shutdown
Jul 18 12:20:31 [main] listen_handler(825) accepted a new connection: 14
Jul 18 12:20:31 [main] client_rx_handler(583) connection from: 14, 127.0.0.1:47189
Jul 18 12:20:31 [main] queue_request(329) GET_NODE_LIST
Jul 18 12:20:31 [io 31] do_process_work(1029) 82, 0 , 1
Jul 18 12:20:31 [main] client_tx_handler(669) connection from: 14, 127.0.0.1:47189
Jul 18 12:20:31 [main] client_handler(770) connection seems to be dead
Jul 18 12:20:31 [main] clear_client(709) refcnt:0, fd:14, 127.0.0.1:47189
Jul 18 12:20:31 [main] destroy_client(678) connection from: 127.0.0.1:47189
Jul 18 12:20:31 [main] listen_handler(825) accepted a new connection: 14
Jul 18 12:20:31 [main] client_rx_handler(583) connection from: 14, 127.0.0.1:47190
Jul 18 12:20:31 [main] queue_request(329) SHUTDOWN
Jul 18 12:20:31 [main] queue_cluster_request(340) SHUTDOWN (0x18b33b0)
Jul 18 12:20:31 [main] cdrv_cpg_deliver(454) 3
Jul 18 12:20:31 [main] sd_notify_handler(854) op SHUTDOWN, size: 96, from: IPv4 ip:172.30.0.80 port:7000
Jul 18 12:20:31 [main] client_tx_handler(669) connection from: 14, 127.0.0.1:47190
Jul 18 12:20:31 [main] main(318) shutdown
Jul 18 12:20:32 [main] log_close(535) logger pid 21227 stopped

Thanks Jens