def _spark_script_args(self, step_num, last_step_num=None):
"""Generate spark harness args for streaming steps (and handle
other spark step types the usual way).
"""
step = self._get_step(step_num)
if step['type'] != 'streaming':
return super(SparkMRJobRunner,
self)._spark_script_args(step_num, last_step_num)
if last_step_num is None:
last_step_num = step_num
args = []
# class name
args.append('%s.%s' %
(self._job_script_module_name(), self._mrjob_cls.__name__))
# INPUT
args.append(','.join(self._step_input_uris(step_num)))
# OUTPUT
# note that we use the output dir for the *last* step
args.append(self._step_output_uri(last_step_num))
# --counter-output-dir, to simulate counters
args.extend(
['--counter-output-dir',
self._counter_output_dir(step_num)])
# --first-step-num, --last-step-num (step range)
args.extend([
'--first-step-num',
str(step_num), '--last-step-num',
str(last_step_num)
])
# --job-args (passthrough args)
job_args = self._mr_job_extra_args()
if job_args:
args.extend(['--job-args', cmd_line(job_args)])
# --compression-codec
jobconf = self._jobconf_for_step(step_num)
compress_conf = jobconf_from_dict(
jobconf, 'mapreduce.output.fileoutputformat.compress')
codec_conf = jobconf_from_dict(
jobconf, 'mapreduce.output.fileoutputformat.compress.codec')
if compress_conf and compress_conf != 'false' and codec_conf:
args.extend(['--compression-codec', codec_conf])
# --num-reducers
num_reducers = jobconf_from_dict(jobconf, 'mapreduce.job.reduces')
if num_reducers and int(num_reducers) > 0:
args.extend(['--num-reducers', str(num_reducers)])
return args
def _invoke_step(self, step_num, step_type):
"""Run the mapper or reducer for the given step.
"""
step = self._get_step(step_num)
if step['type'] != 'streaming':
raise Exception("LocalMRJobRunner cannot run %s steps" %
step['type'])
jobconf = self._jobconf_for_step(step_num)
outfile_prefix = 'step-%04d-%s' % (step_num, step_type)
# allow setting number of tasks from jobconf
if step_type == 'reducer':
num_tasks = int(
jobconf_from_dict(jobconf, 'mapreduce.job.reduces',
self._DEFAULT_REDUCE_TASKS))
else:
num_tasks = int(
jobconf_from_dict(jobconf, 'mapreduce.job.maps',
self._DEFAULT_MAP_TASKS))
# get file splits for mappers and reducers
keep_sorted = (step_type == 'reducer')
file_splits = self._get_file_splits(self._step_input_paths(),
num_tasks,
keep_sorted=keep_sorted)
# since we have grapped the files from the _prev_outfiles as input
# to this step reset _prev_outfiles
self._prev_outfiles = []
# Start the tasks associated with the step:
# if we need to sort, then just sort all input files into one file
# otherwise, split the files needed for mappers and reducers
# and setup the task environment for each
# The correctly-ordered list of task_num, file_name pairs
file_tasks = sorted([(t['task_num'], file_name)
for file_name, t in file_splits.items()],
key=lambda t: t[0])
for task_num, input_path in file_tasks:
# make a new working_dir for each task
working_dir = os.path.join(self._get_local_tmp_dir(),
'job_local_dir', str(step_num),
step_type, str(task_num))
self._setup_working_dir(working_dir)
log.debug("File name %s" % input_path)
# setup environment variables
split_kwargs = {}
if step_type == 'mapper':
# mappers have extra file split info
split_kwargs = dict(
input_file=file_splits[input_path]['orig_name'],
input_start=file_splits[input_path]['start'],
input_length=file_splits[input_path]['length'])
env = self._subprocess_env(step_num, step_type, task_num,
working_dir, **split_kwargs)
output_path = os.path.join(
self._get_local_tmp_dir(),
outfile_prefix + '_part-%05d' % task_num)
log.debug('Writing to %s' % output_path)
self._run_step(step_num, step_type, input_path, output_path,
working_dir, env)
self._prev_outfiles.append(output_path)
self._per_step_runner_finish(step_num)
counters = self._counters[step_num]
if counters:
log.info(_format_counters(counters))
def test_get_missing_jobconf_not_in_table(self):
# there was a bug where defaults didn't work for jobconf
# variables that we don't know about
self.assertEqual(jobconf_from_dict({}, 'user.defined'), None)
self.assertEqual(jobconf_from_dict({}, 'user.defined', 'beauty'),
'beauty')
def test_default(self):
self.assertEqual(jobconf_from_dict({}, 'user.name'), None)
self.assertEqual(jobconf_from_dict({}, 'user.name', 'dave'), 'dave')
def test_get_new_hadoop_jobconf(self):
jobconf = {'mapreduce.job.user.name': 'Edsger W. Dijkstra'}
self.assertEqual(jobconf_from_dict(jobconf, 'user.name'),
'Edsger W. Dijkstra')
self.assertEqual(jobconf_from_dict(jobconf, 'mapreduce.job.user.name'),
'Edsger W. Dijkstra')
def test_get_missing_jobconf_not_in_table(self):
# there was a bug where defaults didn't work for jobconf
# variables that we don't know about
self.assertEqual(jobconf_from_dict({}, 'user.defined'), None)
self.assertEqual(
jobconf_from_dict({}, 'user.defined', 'beauty'), 'beauty')
def test_default(self):
self.assertEqual(jobconf_from_dict({}, 'user.name'), None)
self.assertEqual(jobconf_from_dict({}, 'user.name', 'dave'), 'dave')
def test_get_new_hadoop_jobconf(self):
jobconf = {'mapreduce.job.user.name': 'Edsger W. Dijkstra'}
self.assertEqual(jobconf_from_dict(jobconf, 'user.name'),
'Edsger W. Dijkstra')
self.assertEqual(jobconf_from_dict(jobconf, 'mapreduce.job.user.name'),
'Edsger W. Dijkstra')
def _invoke_step(self, step_num, step_type):
"""Run the mapper or reducer for the given step.
"""
step = self._get_step(step_num)
if step['type'] != 'streaming':
raise Exception("LocalMRJobRunner cannot run %s steps" %
step['type'])
jobconf = self._jobconf_for_step(step_num)
outfile_prefix = 'step-%d-%s' % (step_num, step_type)
# allow setting number of tasks from jobconf
if step_type == 'reducer':
num_tasks = int(jobconf_from_dict(
jobconf, 'mapreduce.job.reduces', self._DEFAULT_REDUCE_TASKS))
else:
num_tasks = int(jobconf_from_dict(
jobconf, 'mapreduce.job.maps', self._DEFAULT_MAP_TASKS))
# get file splits for mappers and reducers
keep_sorted = (step_type == 'reducer')
file_splits = self._get_file_splits(
self._step_input_paths(), num_tasks, keep_sorted=keep_sorted)
# since we have grapped the files from the _prev_outfiles as input
# to this step reset _prev_outfiles
self._prev_outfiles = []
# Start the tasks associated with the step:
# if we need to sort, then just sort all input files into one file
# otherwise, split the files needed for mappers and reducers
# and setup the task environment for each
# The correctly-ordered list of task_num, file_name pairs
file_tasks = sorted([
(t['task_num'], file_name) for file_name, t
in file_splits.items()], key=lambda t: t[0])
for task_num, input_path in file_tasks:
# make a new working_dir for each task
working_dir = os.path.join(
self._get_local_tmp_dir(),
'job_local_dir', str(step_num), step_type, str(task_num))
self._setup_working_dir(working_dir)
log.debug("File name %s" % input_path)
# setup environment variables
split_kwargs = {}
if step_type == 'mapper':
# mappers have extra file split info
split_kwargs = dict(
input_file=file_splits[input_path]['orig_name'],
input_start=file_splits[input_path]['start'],
input_length=file_splits[input_path]['length'])
env = self._subprocess_env(
step_num, step_type, task_num, working_dir, **split_kwargs)
output_path = os.path.join(
self._get_local_tmp_dir(),
outfile_prefix + '_part-%05d' % task_num)
log.info('writing to %s' % output_path)
self._run_step(step_num, step_type, input_path, output_path,
working_dir, env)
self._prev_outfiles.append(output_path)
self.per_step_runner_finish(step_num)
self.print_counters([step_num + 1])
def _spark_script_args(self, step_num, last_step_num=None):
"""Generate spark harness args for streaming steps (and handle
other spark step types the usual way).
"""
if last_step_num is None:
last_step_num = step_num
steps = self._get_steps()[step_num:last_step_num + 1]
if steps[0]['type'] != 'streaming':
return super(SparkMRJobRunner, self)._spark_script_args(
step_num, last_step_num)
args = []
# class name
args.append('%s.%s' % (self._job_script_module_name(),
self._mrjob_cls.__name__))
# INPUT
args.append(
','.join(self._step_input_uris(step_num)))
# OUTPUT
# note that we use the output dir for the *last* step
args.append(
self._step_output_uri(last_step_num))
# --hadoop-input-format. Pass '' to indicate we know there is none
args.extend(['--hadoop-input-format',
self._hadoop_input_format or ''])
# --hadoop-output-format. Pass '' to indicate we know there is none
args.extend(['--hadoop-output-format',
self._hadoop_output_format or ''])
# --sort-values
if self._sort_values:
args.append('--sort-values')
else:
args.append('--no-sort-values')
# --steps-desc
args.extend(['--steps-desc', json.dumps(steps)])
# --counter-output-dir, to simulate counters
args.extend(['--counter-output-dir',
self._counter_output_dir(step_num)])
# --first-step-num, --last-step-num (step range)
args.extend(['--first-step-num', str(step_num),
'--last-step-num', str(last_step_num)])
# --job-args (passthrough args)
# if on local[*] master, keep file upload args as-is (see #2031)
job_args = self._mr_job_extra_args(
local=not self._spark_executors_have_own_wd())
if job_args:
args.extend(['--job-args', cmd_line(job_args)])
# --compression-codec
jobconf = self._jobconf_for_step(step_num)
compress_conf = jobconf_from_dict(
jobconf, 'mapreduce.output.fileoutputformat.compress')
codec_conf = jobconf_from_dict(
jobconf, 'mapreduce.output.fileoutputformat.compress.codec')
if compress_conf and compress_conf != 'false' and codec_conf:
args.extend(['--compression-codec', codec_conf])
# --num-reducers
num_reducers = jobconf_from_dict(jobconf, 'mapreduce.job.reduces')
if num_reducers and int(num_reducers) > 0:
args.extend(['--num-reducers', str(num_reducers)])
# --max-output-files
if self._max_output_files:
args.extend(['--max-output-files',
str(self._max_output_files)])
if self._opts['emulate_map_input_file']:
args.append('--emulate-map-input-file')
return args
def _spark_script_args(self, step_num, last_step_num=None):
"""Generate spark harness args for streaming steps (and handle
other spark step types the usual way).
"""
if last_step_num is None:
last_step_num = step_num
steps = self._get_steps()[step_num:last_step_num + 1]
if steps[0]['type'] != 'streaming':
return super(SparkMRJobRunner, self)._spark_script_args(
step_num, last_step_num)
args = []
# class name
args.append('%s.%s' % (self._job_script_module_name(),
self._mrjob_cls.__name__))
# INPUT
args.append(
','.join(self._step_input_uris(step_num)))
# OUTPUT
# note that we use the output dir for the *last* step
args.append(
self._step_output_uri(last_step_num))
# --hadoop-input-format. Pass '' to indicate we know there is none
args.extend(['--hadoop-input-format',
self._hadoop_input_format or ''])
# --hadoop-output-format. Pass '' to indicate we know there is none
args.extend(['--hadoop-output-format',
self._hadoop_output_format or ''])
# --sort-values
if self._sort_values:
args.append('--sort-values')
else:
args.append('--no-sort-values')
# --steps-desc
args.extend(['--steps-desc', json.dumps(steps)])
# --counter-output-dir, to simulate counters
args.extend(['--counter-output-dir',
self._counter_output_dir(step_num)])
# --first-step-num, --last-step-num (step range)
args.extend(['--first-step-num', str(step_num),
'--last-step-num', str(last_step_num)])
# --job-args (passthrough args)
# if on local[*] master, keep file upload args as-is (see #2031)
job_args = self._mr_job_extra_args(
local=not self._spark_executors_have_own_wd())
if job_args:
args.extend(['--job-args', cmd_line(job_args)])
# --compression-codec
jobconf = self._jobconf_for_step(step_num)
compress_conf = jobconf_from_dict(
jobconf, 'mapreduce.output.fileoutputformat.compress')
codec_conf = jobconf_from_dict(
jobconf, 'mapreduce.output.fileoutputformat.compress.codec')
if compress_conf and compress_conf != 'false' and codec_conf:
args.extend(['--compression-codec', codec_conf])
# --num-reducers
num_reducers = jobconf_from_dict(jobconf, 'mapreduce.job.reduces')
if num_reducers and int(num_reducers) > 0:
args.extend(['--num-reducers', str(num_reducers)])
# --max-output-files
if self._max_output_files:
args.extend(['--max-output-files',
str(self._max_output_files)])
return args
def test_default(self):
self.assertEqual(jobconf_from_dict({}, "user.name"), None)
self.assertEqual(jobconf_from_dict({}, "user.name", "dave"), "dave")
def test_get_new_hadoop_jobconf(self):
jobconf = {"mapreduce.job.user.name": "Edsger W. Dijkstra"}
self.assertEqual(jobconf_from_dict(jobconf, "user.name"), "Edsger W. Dijkstra")
self.assertEqual(jobconf_from_dict(jobconf, "mapreduce.job.user.name"), "Edsger W. Dijkstra")
