After installation, set up issho by following the instructions in setup.

Basic Commands

To use issho in a project:

from issho import Issho

The first thing to do:

devbox = Issho('dev')

This will set up a connection to the machine referred to as dev in your .ssh/config. Note that this will only work if Issho has already been configured.

To run a command on your devbox, you can do the following:

devbox.exec('echo "Hello, world!"')
'Hello, world!'

Note that the data is printed, not returned.

You can copy a file to or from your remote using put & get:

output_filename = 'test.txt'
copy_back_filename = 'get_test.txt'
with open(output_filename, 'w') as f:
    f.write('\n'.join(map(str, range(5))))
devbox.exec('cat {}'.format(output_filename))
devbox.get(output_filname, copy_back_filename)
for line in open(copy_back_filename):

Convenience Functions

Shell Commands

Instead of using devbox.exec(cmd, *args), you can write devbox.cmd(*args):

devbox.touch('my_test.txt')' | grep my_test.txt')

Underscores in the function name are converted to spaces:


Hadoop & HDFS

Hadoop functions can be accessed using the .hadoop or .hdfs methods. You do not need to prepend the dash to hadoop operations, though they will still work with it:

devbox.hdfs('ls /tmp | grep test')
devbox.hadoop('mkdir -p /tmp/test/')

put and get can also get from HDFS, if passed a qualified HDFS path, or if the hadoop option is passed.:

devbox.put('test.txt', '/tmp/my_folder/', hadoop=True)


issho offers several convenience functions, including this for Hive:

devbox.hive('select * from burgers limit 10;')

Results from hive queries can be output locally by passing an output_filename:

devbox.hive('select stack(3, "hello", "cruel", "world") as val;', "hello.tsv")


issho can trigger a spark job using spark-submit; you can call it using `spark_submit or spark:

devbox.spark(application='test.jar', application_class='com.test.SparkWorkflow'...)