python子进程输出被截断

Question

While running shell command in python using subprocess I am getting wrong output.

grep ( com.vertica.solutions.kafka.Launcher ) in bash shell

ps -ef | grep com.vertica.solutions.kafka.Launcher
root      92300  39024  0 23:06 pts/1    00:00:00 grep --color=auto  com.vertica.solutions.kafka.Launcher
dbadmin  413872 413868  0 22:06 pts/0    00:00:24 java -cp /opt/vertica/packages/kafka/bin/../lib/*:/opt/vertica/java/lib/vertica-jdbc.jar:/opt/vertica/packages/kafka/bin/../config/* -Dlog4j.configurationFile=/opt/vertica/packages/kafka/bin/../config/vkafka-log-config.xml -DLog4jContextSelector=org.apache.logging.log4j.core.async.AsyncLoggerContextSelector com.vertica.solutions.kafka.Launcher --conf /home/dbadmin/live.conf

grep ( com.vertica.solutions.kafka.Launcher ) in python shell

>>> cmd = "ps aux |grep com.vertica.solutions.kafka.Launcher| grep -v grep"
>>> check_output(cmd,shell=True).strip()
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/lib64/python2.7/subprocess.py", line 575, in check_output
raise CalledProcessError(retcode, cmd, output=output)
subprocess.CalledProcessError: Command 'ps aux |grep com.vertica.solutions.kafka.Launcher| grep -v grep' returned non-zero exit status 1

This is because subprocess is truncating ps output. As when I grep with word kafka I get the result but its not desired.

grep only kafka

>>> cmd = "ps aux | grep kafka | grep -v grep"
>>> check_output(cmd,shell=True).strip()
dbadmin  413868  0.0  0.0 113128  1372 pts/0    S    22:06   0:00   /bin/bash /opt/vertica/packages/kafka/bin/vkconfig launch --conf /home/dbadmin/live.conf\n
dbadmin  413872  0.6  1.0 7403000 175544 pts/0  Sl   22:06   0:25 java -cp /opt/vertica/packages/kafka/bin/../lib/*:/opt/vertica/java/lib/vertica-jdbc.jar:/opt/vertica/package'

When I grep only with word kafka I get the process (pid 413872) but in python shell COMMAND column output of process command is incomplete.

Answer 1

Subprocess isn't truncating whatever get's written into stdout , it is piped to your python session. What happens is that -v isn't successful so it exits with a bad return code (None Zero).

To ignore the cases where you don't get any output from grep you can do the following:

import os
from subprocess import Popen

DEVNULL = open(os.devnull, 'wb')
cmd = 'ps aux |grep com.vertica.solutions.kafka.Launcher| grep -v grep' 
process = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=DEVNULL, shell=True)
print process.communicate()[0] #output 

That will make sure, in case the grep isn't successful, instead of returning an error code, it will just dump it.