Panda Song created MAPREDUCE-7209:
-------------------------------------
Summary: getting error "Can not create a Path from an empty
string" when running mapreduce job
Key: MAPREDUCE-7209
URL: https://issues.apache.org/jira/browse/MAPREDUCE-7209
Project: Hadoop Map/Reduce
Issue Type: Bug
Reporter: Panda Song
hdfs:3.1.1
Mapreduce came into the error "Can not create a Path from an empty string" ,It
just happened with the specific input file
The input file is too large to upload,can somebody help me ?
{code:java}
#!/usr/bin/python3
from mrjob.job import MRJob
from mrjob.step import MRStep
import mrjob.protocol as protocol
import sys
import json
class QM(MRJob):
FILES = ['utils.py', 'models.py', 'stats.py']
LIBJARS = ['nicknack-1.0.1.jar']
JOBCONF = {
'mapred.reduce.tasks': 4,
'mapreduce.output.fileoutputformat.compress': "true",
'mapreduce.output.fileoutputformat.compress.codec':
'org.apache.hadoop.io.compress.GzipCodec'
}
INPUT_PROTOCOL = protocol.TextValueProtocol
INTERNAL_PROTOCOL = protocol.JSONProtocol
OUTPUT_PROTOCOL = protocol.RawProtocol
HADOOP_OUTPUT_FORMAT = 'nicknack.MultipleValueOutputFormat'
def mapper(self, _, value):
from models import OpenApiOpen, ThirdPart_UserOpenApp,
filter_baidu_pan_android, filter_vivo
from stats import Stats
from utils import get_mr_input_file_name, DATA
try:
input_file_name = get_mr_input_file_name()
if 'ThirdPart_UserOpenApp' in input_file_name:
r = ThirdPart_UserOpenApp.parse_qm(value)
elif 'OpenApiOpen' in input_file_name:
r = OpenApiOpen.parse_qm(value)
else:
r = None
if r is None:
return Stats.on_error(self)
ok, client_id, did, imei, v = r
yield DATA, v
else:
return Stats.on_error(self)
except Exception as e:
return Stats.on_exception(self)
if __name__ == '__main__':
QM.run()
{code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: mapreduce-dev-unsubscr...@hadoop.apache.org
For additional commands, e-mail: mapreduce-dev-h...@hadoop.apache.org
