3 In general, signals need only be sent to the master process. However,
4 the signals unicorn uses internally to communicate with the worker
5 processes are documented here as well.
9 * HUP - reload config file, app, and gracefully restart all workers
11 * INT/TERM - quick shutdown, kills all workers immediately
13 * QUIT - graceful shutdown, waits for workers to finish their
14 current request before finishing.
16 * USR1 - reopen all logs owned by the master and all workers
17 See Unicorn::Util.reopen_logs for what is considered a log.
19 * USR2 - reexecute the running binary. A separate QUIT
20 should be sent to the original process once the child is verified to
23 * WINCH - gracefully stops workers but keep the master running.
24 This will only work for daemonized processes.
26 * TTIN - increment the number of worker processes by one
28 * TTOU - decrement the number of worker processes by one
32 Sending signals directly to the worker processes should not normally be
33 needed. If the master process is running, any exited worker will be
34 automatically respawned.
36 * INT/TERM - Quick shutdown, immediately exit.
37 Unless WINCH has been sent to the master (or the master is killed),
38 the master process will respawn a worker to replace this one.
40 * QUIT - Gracefully exit after finishing the current request.
41 Unless WINCH has been sent to the master (or the master is killed),
42 the master process will respawn a worker to replace this one.
44 * USR1 - Reopen all logs owned by the worker process.
45 See Unicorn::Util.reopen_logs for what is considered a log.
46 Log files are not reopened until it is done processing
47 the current request, so multiple log lines for one request
48 (as done by Rails) will not be split across multiple logs.
50 === Procedure to replace a running unicorn executable
52 You may replace a running instance of unicorn with a new one without
53 losing any incoming connections. Doing so will reload all of your
54 application code, Unicorn config, Ruby executable, and all libraries.
55 The only things that will not change (due to OS limitations) are:
57 1. The listener backlog size of already-bound sockets
59 2. The path to the unicorn executable script. If you want to change to
60 a different installation of Ruby, you can modify the shebang
61 line to point to your alternative interpreter.
63 The procedure is exactly like that of nginx:
65 1. Send USR2 to the master process
67 2. Check your process manager or pid files to see if a new master spawned
68 successfully. If you're using a pid file, the old process will have
69 ".oldbin" appended to its path. You should have two master instances
70 of unicorn running now, both of which will have workers servicing
71 requests. Your process tree should look something like this:
84 3. You can now send WINCH to the old master process so only the new workers
85 serve requests. If your unicorn process is bound to an interactive
86 terminal, you can skip this step. Step 5 will be more difficult but
87 you can also skip it if your process is not daemonized.
89 4. You should now ensure that everything is running correctly with the
90 new workers as the old workers die off.
92 5. If everything seems ok, then send QUIT to the old master. You're done!
94 If something is broken, then send HUP to the old master to reload
95 the config and restart its workers. Then send QUIT to the new master