Distributed processing of a batch of tasks.

TL;DR

Create a file Tasks with a list of commands you want to run:

myprog arg0 &> myprog_0.log
myprog arg1 &> myprog_1.log
...
myprog argN &> myprog_N.log

Then run:

disBatch -s localhost:5 Tasks

to execute all of the command invocations in Tasks. disBatch will start the first five running concurrently on your local machine. When one finishes, the next will be started until all are done.

Or run:

sbatch -n 5 disBatch Tasks

to execute the tasks using computational resources allocated via SLURM. (Of course, you may need to provide additional SLURM arguments to specify a partition, time limit, etc, and see the discussion about PYTHONPATH.)

Overview

One common usage pattern for distributed computing involves processing a long list of commands (aka tasks):

( cd /path/to/workdir ; source SetupEnv ; myprog -a 0 -b 0 -c 0 ) &> task_0_0_0.log
( cd /path/to/workdir ; source SetupEnv ; myprog -a 0 -b 0 -c 1 ) &> task_0_0_1.log
...
( cd /path/to/workdir ; source SetupEnv ; myprog -a 9 -b 9 -c 8 ) &> task_9_9_8.log
( cd /path/to/workdir ; source SetupEnv ; myprog -a 9 -b 9 -c 9 ) &> task_9_9_9.log

One could do this by submitting 1,000 separate jobs to a cluster computer, but that may present problems for the queuing system and can behave badly if the system is configured to handle jobs in a simple first come, first serve fashion.

Another alternative is to use a resource management system's native support for job arrays, but this often requires massaging the commands to reflect syntax specific to a particular system's implementation of job arrays.

And what if you don't have a cluster available, but do have a collection of networked computers? Or you just want to make use of multiple cores on your own computer?

In any event, when processing such a list of tasks, it is helpful to acquire metadata about the execution of each task: where it ran, how long it took, its exit return code, etc.

disBatch has been designed to support this usage in a simple and portable way, as well as to provide the sort of metadata that can be helpful for debugging and reissuing failed tasks.

It can take as input a file, each of whose lines is a task in the form of a command sequence. For example, the file could consists of the 1000 commands listed above. It launches the tasks one after the other until all specified execution resources are in use. Then as one executing task exits, the next task in the file is launched. This repeats until all the lines in the file have been processed.

Each task is run in a new shell. If you want to manipulate the execution environment of a task, add the appropriate operations to the command sequence—source SetupEnv in the above is one example. For another, if you just need to set an environment variable, each task line would look something like:

export LD_LIBRARY_PATH=/d0/d1/d2:$LD_LIBRARY_PATH ; rest ; of ; command ; sequence

Or, for more complex set ups, command sequences and input/output redirection requirements, you could place everything in a small shell script with appropriate arguments for the parts that vary from task to task, say RunMyprog.sh:

#!/bin/bash

id=$1
shift
cd /path/to/workdir
module purge
module load gcc openblas python3 

export LD_LIBRARY_PATH=/d0/d1/d2:$LD_LIBRARY_PATH
myProg "$@" > results/${id}.out 2> logs/${id}.log

The task file would then contain:

./RunMyprog.sh 0_0_0 -a 0 -b 0 -c 0
./RunMyprog.sh 0_0_1 -a 0 -b 0 -c 1
...
./RunMyprog.sh 9_9_8 -a 9 -b 9 -c 8
./RunMyprog.sh 9_9_9 -a 9 -b 9 -c 9

See #DISBATCH directives for ways to simplify task lines. DisBatch also sets some environment variables that can be used in your commands as arguments or to generate task-specifc file names:

• DISBATCH_JOBID: A name disBatch creates that should be unique to the job
• DISBATCH_NAMETASKS: The basename of the task file
• DISBATCH_REPEAT_INDEX: See the repeat construct in #DISBATCH directives
• DISBATCH_STREAM_INDEX: The 1-based line number of the line from the task file that generated the task " DISBATCH_TASKID: 0-based sequential counter value that uniquely identifies each task

Appending _ZP to any of the last three will produce a 0-padded value (to six places). If these variables are used to create file names, 0-padding will result in files names that sort correctly.

Once you have created the task file, running disBatch is straightforward. For example, working with a cluster managed by SLURM, all that needs to be done is to submit a job like the following:

sbatch -n 20 -c 4 disBatch TaskFileName

This particular invocation will allocate sufficient resources to process 20 tasks at a time, each of which needs 4 cores. disBatch will use environment variables initialized by SLURM to determine the execution resources to use for the run. This invocation assumes an appropriately installed disBatch is in your PATH, see installation for details.

disBatch also allows the pool of execution resources to be increased or decreased during the course of a run:

sbatch -n 10 -c 4 ./TaskFileName_dbUtil.sh

will add enough resources to run 10 more tasks concurrently. TaskFileName_dbUtl.sh is a utility script created by disBatch when the run starts (the actual name is a little more complex, see startup).

Various log files will be created as the run unfolds:

• TaskFileName_*_status.txt: status of every task (details below). * elides a unique identifier disBatch creates to distinguish one run from another
• TaskFileName_*_[context|driver|engine].log: The disBatch driver log file contains details mostly of interest in case of a problem with disBatch itself. (The driver log file name can be changed with -l). It can generally be ignored by end users (but keep it around in the event that something did go wrong—it will aid debugging). The *_[context|engine].log files contain similar information for the disBatch components that manage execution resources.
• disBatch_*_kvsinfo.txt: TCP address of invoked KVS server if any (for additional advanced status monitoring)

While disBatch is a python3 application, it can run tasks from any language environment—anything you can run from a shell, can be run as a task.

Status file

The _status.txt file contains tab-delimited lines of the form:

314	315	-1	worker032	8016	0	10.0486528873	1458660919.78	1458660929.83	0	""	0	""	cd /path/to/workdir ; myprog -a 3 -b 1 -c 4 > task_3_1_4.log 2>&1

These fields are:

1. Flags: The first field, blank in this case, may contain E, O, R, B, or S flags. Each program/task should be invoked in such a way that standard error and standard output end up in appropriate files. If that is not the case E or O flags will be raised. R indicates that the task returned a non-zero exit code. B indicates a barrier. S indicates the job was skipped (this may happen during "resume" runs).
2. Task ID: The 314 is the 0-based index of the task (starting from the beginning of the task file, incremented for each task, including repeats).
3. Line number: The 315 is the 1-based line from the task file. Blank lines, comments, directives and repeats may cause this to drift considerably from the value of Task ID.
4. Repeat index: The -1 is the repeat index (as in this example, -1 indicates this task was not part of a repeat directive).
5. Node: worker032 identifies the node on which the task ran.
6. PID: 8016 is the PID of the bash shell used to run the task.
7. Exit code: 0 is the exit code returned.
8. Elapsed time: 10.0486528873 (seconds),
9. Start time:1458660919.78 (epoch based),
10. Finish time: 1458660929.83 (epoch based).
11. Bytes of leaked output (not redirected to a file),
12. Output snippet (up to 80 bytes consisting of the prefix and suffix of the output),
13. Bytes of leaked error output,
14. Error snippet,
15. Command: cd ... is the text of the task (repeated from the task file, but subject to modification by directives).

Installation

Users of Flatiron resources: disBatch is available via the module system. You do not need to clone this repo to use it.

disBatch requires the kvsstcp package, which should be installed in python's path, or placed in this directory. You can simply clone this git repository with --recursive (or run git submodule update --init if you've already cloned it) to get both. No additional installation steps are required, but you can use pip to add it to a particular python installation.

You can run disBatch directly from the top level directory of the git clone, but depending on your execution environment (e.g., when using SLURM), the ability of disBatch to determine the location of itself and kvsstcp may be disrupted. To avoid such problems, set the environment variable PYTHONPATH to include the path of the git clone (the directory containing this "Readme.md" file) or install with pip and invoke disBatch from that installation in your SLURM submission.

Execution Environments

disBatch is designed to support a variety of execution environments, from your own desktop, to a local collection of workstations, to large clusters managed by job schedulers. It currently supports SLURM and can be executed from sbatch, but it is architected to make it simple to add support for other resource managers.

You can also run directly on one or more machines by setting an environment variable:

DISBATCH_SSH_NODELIST=localhost:7,otherhost:3

or specifying an invocation argument:

-s localhost:7,otherhost:3

This allows execution directly on your localhost and via ssh for remote hosts without the need for a resource management system. In this example, disBatch is told it can use seven CPUs on your local host and three on otherhost. Assuming the default mapping of one task to one CPU applies in this example, seven tasks could be in progress at any given time on localhost, and three on otherhost. Note that localhost is an actual name you can use to refer to the machine on which you are currently working. otherhost is fictious. Hosts used via ssh must be set up to allow ssh to work without a password and must share the working directory for the disBatch run.

disBatch refers to a collection of execution resources as a context and the resources proper as engines. So the SLURM example sbatch -n 20 -c 4, run on a cluster with 16-core nodes, might create one context with five engines (one each for five 16-core nodes, capable of running four concurrent 4-core tasks each), while the SSH example creates one context with two engines (capable of running seven and three concurrent tasks, respectively).

Invocation

usage: disBatch [-h] [-e] [--force-resume] [--kvsserver [HOST:PORT]]
                [--logfile FILE] [--mailFreq N] [--mailTo ADDR] [-p PATH]
                [-r STATUSFILE] [-R] [-S] [--status-header] [--use-address HOST:PORT]
                [-w] [--taskcommand COMMAND] [--taskserver [HOST:PORT]]
                [-C TASK_LIMIT] [-c N] [--fill] [-g] [--no-retire]
                [-l COMMAND] [--retire-cmd COMMAND] [-s HOST:COUNT] [-t N]
                [taskfile]

Use batch resources to process a file of tasks, one task per line.

positional arguments:
  taskfile              File with tasks, one task per line ("-" for stdin)

optional arguments:
  -h, --help            show this help message and exit
  -e, --exit-code       When any task fails, exit with non-zero status
                        (default: only if disBatch itself fails)
  --force-resume        With -r, proceed even if task commands/lines are
                        different.
  --kvsserver [HOST:PORT]
                        Use a running KVS server.
  --logfile FILE        Log file.
  --mailFreq N          Send email every N task completions (default: 1). "--
                        mailTo" must be given.
  --mailTo ADDR         Mail address for task completion notification(s).
  -p PATH, --prefix PATH
                        Path for log, dbUtil, and status files (default: ".").
                        If ends with non-directory component, use as prefix
                        for these files names (default:
			  <Taskfile>_disBatch_<YYYYMMDDhhmmss>_<Random>
			).
  -r STATUSFILE, --resume-from STATUSFILE
                        Read the status file from a previous run and skip any
                        completed tasks (may be specified multiple times).
  -R, --retry           With -r, also retry any tasks which failed in previous
                        runs (non-zero return).
  -S, --startup-only    Startup only the disBatch server (and KVS server if
                        appropriate). Use "dbUtil..." script to add execution
                        contexts. Incompatible with "--ssh-node".
  --status-header       Add header line to status file.
  --use-address HOST:PORT
                        Specify hostname and port to use for this run.
  -w, --web             Enable web interface.
  --taskcommand COMMAND
                        Tasks will come from the command specified via the KVS
                        server (passed in the environment).
  --taskserver [HOST:PORT]
                        Tasks will come from the KVS server.
context specific arguments:
  -C TASK_LIMIT, --context-task-limit TASK_LIMIT
                        Shutdown after running COUNT tasks (0 => no limit).
  -c N, --cpusPerTask N
                        Number of cores used per task; may be fractional
                        (default: 1).
  --fill                Try to use extra cores if allocated cores exceeds
                        requested cores.
  -g, --gpu             Use assigned GPU resources
  --no-retire           Don't retire nodes from the batch system (e.g., if
                        running as part of a larger job); equivalent to -k ''.
  -l COMMAND, --label COMMAND
                        Label for this context. Should be unique.
  --retire-cmd COMMAND  Shell command to run to retire a node (environment
                        includes $NODE being retired, remaining $ACTIVE node
                        list, $RETIRED node list; default based on batch
                        system). Incompatible with "--ssh-node".
  -s HOST:COUNT, --ssh-node HOST:COUNT
                        Run tasks over SSH on the given nodes (can be
                        specified multiple times for additional hosts;
                        equivalent to setting DISBATCH_SSH_NODELIST)
  -t N, --tasksPerNode N
                        Maximum concurrently executing tasks per node (up to
                        cores/cpusPerTask).

The options for mail will only work if your computing environment permits processes to access mail via SMTP.

A value for -c < 1 effectively allows you to run more tasks concurrently than CPUs specified for the run. This is somewhat unusual, and generally not recommended, but could be appropriate in some cases.

The --no-retire and --retire-cmd flags allow you to control what disBatch does when a node is no longer needed to run jobs. When running under slurm, disBatch will by default run the command:

scontrol update JobId="$SLURM_JOBID" NodeList="${DRIVER_NODE:+$DRIVER_NODE,}$ACTIVE"

which will tell slurm to release any nodes no longer being used. You can set this to run a different command, or nothing at all. While running this command, the follow environment variables will be set: NODE (the node that is no longer needed), ACTIVE (a comma-delimited list of nodes that are still active), RETIRED (a comma-delimited list of nodes that are no longer active, including $NODE), and possibly DRIVER_NODE (the node still running the main disBatch script, if it's not in ACTIVE).

The -g argument parses the CUDA environment varables (CUDA_VISIBLE_DEVICES, GPU_DEVICE_ORDINAL) provided on each node and divides the resources between the running tasks. For example, with SLURM, if you want to run 12 tasks concurrently, each using 3 CPUs and 1 GPU (that is, 12*3 CPUs and 12 GPUs total), you can do:

sbatch -n 12 -c 3 --gpus-per-task=1 -p gpu disBatch -g TaskFile

-S Startup only mode. In this mode, disBatch starts up the task management system and then waits for execution resources to be added. At startup, disBatch always generates a script <Prefix>_dbUtil.sh, where <Prefix> refers to the -p option or default, see above. We'll call this simply dbUtils.sh here, but remember to include <Prefix>_ in actual use. You can add execution resources by doing one or more of the following multiple times:

1. Submit dbUtils.sh as a job, e.g.:

   sbatch -n 40 dbUtil.sh

2. Use ssh, e.g.:

   ./dbUtil.sh -s localhost:4,friendlyNeighbor:5

Each of these creates an execution context, which contains one of more execution engines (if using, for example, 8-core nodes, then five for the first; two in the second). An engine can run one or more tasks currently. In the first example, each of the five engines will run up to eight tasks concurrently, while in the second example, the engine on localhost will run up to four tasks concurrently and the engine on friendlyNeighbor will run up to five. ./dbUtil.sh --mon will start a simple ASCII-based monitor that tracks the overall state of the disBatch run, and the activity of the individual contexts and engines. By cursoring over an engine, you can send a shutdown signal to the engine or its context. This signal is soft, triggering a graceful shutdown that will occur only after currently assigned tasks are complete. Other execution resources are uneffected.

When a context is started, you can also supply the argument --context-task-limit N. This will shutdown the context and all associated engines after it has run N tasks.

Taken together, these mechanisms enable disBatch to run on a dynamic pool of execution resources, so you can "borrow" a colleague's workstation overnight, or claim a large chunk of a currently idle partition, but return some if demands picks up, or chain together a series of time limited allocations to accomplish a long run. When using this mode, keep in mind two caveats: (i) The time quantum is determined by your task duration. If any given task might run for hours or days, then the utility of this is limited. You can still use standard means (kill, scancel) to terminate contexts and engines, but you will likely have incomplete tasks to reckon with; (ii) The task manangement system must itself be run in a setting where a long lived process is OK. Say in a screen or tmux session on the login node of a cluster, or on your personal workstation (assuming it has the appropriate connectivity to reach the other resources you plan to use).

-r uses the status file of a previous run to determine what tasks to run during this disBatch invocation. Only those tasks that haven't yet run (or with -R, those that haven't run or did but returned a non-0 exit code) are run this time. By default, the numeric task identifier and the text of the command are used to determine if a current task is the same as one found in the status file. --force-resume restricts the comparison to just the numeric identifier.

--use-address HOST:PORT can be used if disBatch is not able to determine the correct hostname for the machine it is running on (or you need to override what was detected). This is often the case when running on a personal laptop without a "real" network configuration. In this case --use-address=localhost:0 will generally be sufficient.

--kvsserver, --taskcommand, and --taskserver implement advanced functionality (placing disBatch in an existing shared key store context and allowing for a programmatic rather than textual task interface). Contact the authors for more details.

Considerations for large runs

If you do submit jobs with order 10000 or more tasks, you should carefully consider how you want to organize the output (and error) files produced by each of the tasks. It is generally a bad idea to have more than a few thousand files in any one directory, so you will probably want to introduce at least one extra level of directory hierarchy so that the files can be divided into smaller groups. Intermediate directory 13, say, might hold all the files for tasks 13000 to 13999.

#DISBATCH directives

PREFIX and SUFFIX

In order to simplify task files, disBatch supports a couple of directives to specify common task prefix strings and suffix strings. As noted above, it also sets environment variables to identify various aspects of the submission. Here's an example

# Note there is a space at the end of the next line.
#DISBATCH PREFIX ( cd /path/to/workdir ; source SetupEnv ; 
#DISBATCH SUFFIX  ) &> ${DISBATCH_NAMETASKS}_${DISBATCH_JOBID}_${DISBATCH_TASKID_ZP}.log

These are textually prepended and appended, respectively, to the text of each subsequent task line. If the suffix includes redirection and a task is a proper command sequence (a series of commands joined by ;), then the task should be wrapped in ( ... ), as in this example, so that the standard error and standard output of the whole sequence will be redirected to the log file. If this is not done, only standard error and standard output for the last component of the command sequence will be captured. This is probably not what you want unless you have redirected these outputs for the previous individual parts of the command sequence.

Using these, the above commands could be replaced with:

myprog -a 0 -b 0 -c 0
myprog -a 0 -b 0 -c 1
... 
myprog -a 9 -b 9 -c 8
myprog -a 9 -b 9 -c 9

Note: the log files will have a different naming scheme, but there will still be one per task.

Later occurrences of #DISBATCH PREFIX or #DISBATCH SUFFIX in a task file simply replace previous ones. When these are used, the tasks reported in the status file include the prefix and suffix in force at the time the task was launched.

BARRIER

If your tasks fall into groups where a later group should only begin after all tasks of the previous group have completely finished, you can use this directive:

#DISBATCH BARRIER

When disBatch encounters this directive, it will not launch another task until all tasks in progress have completed. The following form:

#DISBATCH BARRIER CHECK

checks the exit status of the tasks done since the last barrier (or start of the run). If any task had a non-zero exit status, the run will exit once this barrier is met.

REPEAT

For those problems that are easily handled via a job-array-like approach:

 #DISBATCH REPEAT 5 start 100 step 50 [command]

will expand into five tasks, each with the environment variable DISBATCH_REPEAT_INDEX set to one of 100, 150, 200, 250, or 300. The tasks will consist of the concatenation of the prefix, command (if provided), and the suffix currently in effect. start defaults to 0, step to 1. Note: the semantics here differ somewhat from many range constructs, the number immediately following REPEAT sets the number of tasks that will be executed; the next two numbers affect only the value that the repeat index will have in the environment for each of the repeat task instances. So, returning to our earlier example, the task file could be:

#DISBATCH PREFIX a=$((DISBATCH_REPEAT_INDEX/100)) b=$(((DISBATCH_REPEAT_INDEX%100)/10 )) c=$((DISBATCH_REPEAT_INDEX%10) ; ( cd /path/to/workdir ; source SetupEnv ; myprog -a $a -b $b -c $c ) &> task_${a}_${b}_${c}.log
#DISBATCH REPEAT 1000

This is not a model of clarity, but it does illustrate that the repeat constuct can be relatively powerful. Many users may find it more convenient to use the tool of their choice to generate a text file with 1000 invocations explictly written out.

PERENGINE

#DISBATCH PERENGINE START { command ; sequence ; } &> engine_start_${DISBATCH_ENGINE_RANK}.log
#DISBATCH PERENGINE STOP { command ; sequence ; } &> engine_stop_${DISBATCH_ENGINE_RANK}.log

Use these to specify commands that should run at the time an engine joins a disBatch run or at the time the engine leaves the disBatch run, respectively. You could, for example, use these to bulk copy some heavily referenced read-only data to the engine's local storage area before any tasks are run, and then delete that data when the engine shuts down. You can use the environment variable DISBATCH_ENGINE_RANK to distinguish one engine from another; for example, it is used here to keep log files separate.

These directives must come before any other tasks.

Embedded disBatch

You can start disBatch from within a python script by instantiating a "DisBatcher" object.

See exampleTaskFiles/dberTest.py for an example.

The "DisBatcher" class (defined in disbatch/disBatch.py) illustrates how to interact with disBatch via KVS. This approach could be used to enable similar functionality in other language settings.

License

Copyright 2017 Simons Foundation

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.