1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
|
>>> ms -h
usage: ms [messenger-flags] [task-spooler-flags] [command]
messenger-flags: [--cd directory] [--env "FLAG1=VALUE1;FLAG2=VALUE2;..."] [--host/-H host_name]
[--show_gpus] [--show_free_gpus] [--num_free_gpus] [--auto_server]
[--kill] [--sync directory] [--sync_dest directory]
[--excludes pattern1,pattern2,...]
task-spooler-flags: [-h] [--set_gpu_wait seconds] [--get_gpu_wait] [--get_label]
[--count_running] [--last_queue_id] [--gpus num] [--full_cmd job_id]
[-K] [-C] [-l] [-S num] [-t job_id] [-c job_id] [-p job_id] [-o job_id]
[-i job_id] [-s job_id] [-r job_id] [-w job_id] [-k job_id] [-T]
[-u job_id] [-U job_id1-job_id2] [-B] [-V] [-n] [-E] [-g] [-f] [-m] [-d]
[-D job_id] [-L label] [-N num]
Messenger - A multi-server plugin for Task Spooler
positional arguments:
command job to be run
optional arguments:
-h, --help show this help message and exit
--cd directory change directory. For e.g., ``ms --cd /home/justanhduc/Documents``.
--env FLAG1=VALUE1:FLAG2=VALUE2:...
set environment variable flags.
--host host_num, -H host_num
host to select. Value corresponds to the order specified in the ".servers_ports" file.
--show_gpus show all GPUs info.
--show_free_gpus show current available GPUs info.
--num_free_gpus show the number of available GPUs.
--auto_server auto-magically choose a server based on the number of available GPUs.
--kill kill Messenger server.
--sync directory whether to sync the selected working directory to a temp directory before executing the
command.
--excludes pattern1,pattern2,...
exception patterns when moving files to server.
--set_gpu_wait seconds
set time to wait before running the next GPU job (30 seconds by default)
--get_gpu_wait get time to wait before running the next GPU job.
--get_label, -a show the job label. Of the last added, if not specified.
--count_running, -R return the number of running jobs
--last_queue_id, -q show the job ID of the last added.
--full_cmd job_id show full command. Of the last added, if not specified.
-K kill the task spooler server
-C clear the list of finished jobs
-l show the job list (default action)
-S num get/set the number of max simultaneous jobs of the server.
-t job_id "tail -n 10 -f" the output of the job. Last run if -1.
-c job_id like -t, but shows all the lines. Last run if -1.
-p job_id show the pid of the job. Last run if -1.
-o job_id show the output file. Of last job run, if -1.
-i job_id show job information. Of last job run, if -1.
-s job_id show the job state. Of the last added, if -1.
-r job_id remove a job. The last added, if -1.
-w job_id wait for a job. The last added, if -1.
-k job_id send SIGTERM to the job process group. The last run, if -1.
-T send SIGTERM to all running job groups.
-u job_id put that job first. The last added, if not specified.
-U job_id1-job_id2 swap two jobs in the queue.
-B in case of full queue on the server, quit (2) instead of waiting.
-V show the program version
-n don't store the output of the command.
-E Keep stderr apart, in a name like the output file, but adding '.e'.
-z gzip the stored output (if not -n).
-f don't fork into background.
-m send the output by e-mail (uses sendmail).
-d the job will be run after the last job ends.
-D job_id the job will be run after the job of given id ends.
-L label name this task with a label, to be distinguished on listing.
-N num number of slots required by the job (1 default).
--gpus num, -G num number of GPUs required by the job (1 default).
|