Skip to content

Latest commit

 

History

History
124 lines (106 loc) · 5.05 KB

README.md

File metadata and controls

124 lines (106 loc) · 5.05 KB

pycdap

Python library and CLI for exporting CDAP Pipelines

If you want use the pycdap library in your own python projects simply import the Pipeline module and configure an object instance.

There is also a command line interface (CLI) bundled with this library that lets you use the Pipeline module from the command line for interacting with your CDAP instance.

Setup

To install pycdap run the setup.py installer from within the pycdap directory. The installer will create an executable named piper.

pip install .
-- or --
python setup.py install

piper allows you to interact with your CDAP instance via the console.

CLI Usage

To list the available sub-commands for piper use the --help argument.
Each sub command has usage information available.

piper --help
piper export --help

List Pipelines

Print out a listing of all the available namespaces and pipelines contained within.

piper list

Status of CDAP Instance

To get the version, status and the namespaces available on a CDAP instance, execute:

piper status
-- or --
piper status -u http://10.30.20.123:11015

Export Pipelines

There are two export options available export and export_all.
The former option takes key/word arguments, whereas the later exports all namespaces and all the pipelines per namespace.

piper export -ns default -t draft -o /tmp/my_pipelines
piper export -u http://10.30.20.123:11015

Arguments and Options

export:

  • -u, --cdap_instance ... CDAP instance to connect to. Defaults to http://localhost:11015
  • -ns, --namespace ...... Namespace to export from. If not specified, the default namespace will be used.
  • -t, --type .................... Pipeline to export. Default is to export all pipelines.
  • -o', --output_dir ......... Directory to export the pipelines to. Defaults to the current directory.

export_all:

  • -u, --cdap_instance ... CDAP instance to connect to. Defaults to http://localhost:11015

Usage Examples:

piper export -ns NS1                              #export all pipeline types from the `NS1` namespace 
piper export -u http://10.30.20.123:11015 -t all  #export all pipeline types from the `default` namespace on a remote CDAP instance.
piper export -ns NS1 -t drafts -o ~/cdap/drafts   #export draft pipeline from the `NS1 namespace and write output to /home/username/cdap/drafts
#
piper export_all                                  #export all pipeline from all namespaces 
piper export_all -u http://10.30.20.123:11015     #export all pipeline from all namespaces on a remote CDAP instance.

Library Usage:

import Pipeline
p = Pipeline('http://localhost:11015')
p.connect()
p.list()

LIST

Prints out a listing of all the pipelines

p.list()

APPS = Deployed Pipelines

apps = c.apps()
print json.dumps(p.apps(), indent=2)                  # return all apps in the all namespaces
print json.dumps(p.apps('default'), indent=2)         # return all apps in the 'default' ns
print json.dumps(p.apps('default', 'NS1'), indent=2)  # return all apps in the 'default' and 'NS1' ns
print json.dumps(p.apps('foo'), indent=2)             # Terminate, 'foo' is not a valid namespace
print json.dumps(p.apps('foo', 'default'), indent=2)  # Terminate: even though 'default is valid 'foo' is not

DRAFTS - Draft Pipelines

print json.dumps(p.drafts(), indent=2)                  # return all drafts in all ns
print json.dumps(p.drafts('default'), indent=2)         # return all drafts in 'default' ns
print json.dumps(p.drafts('default', 'NS1'), indent=2)  # return all drafts in 'default' ns
print json.dumps(p.drafts('foo'), indent=2)             # Terminate, 'foo' is not a valid namespace
print json.dumps(p.drafts('default', 'foo'), indent=2)  # Terminate: even though 'default is valid 'foo' is not
print json.dumps(p.drafts('default','default','default'), indent=2)         # return all drafts in 'default' ns

EXPORT - Save Pipelines Configs to Disk

Unless specified, all pipelines will be saved to the current working directory where piper is executed from.

### Parameter options ###
# namespaces = 'n', 'ns', 'namespace'
# pipeline types = 'p', 'pipeline', 'pipelines', 'type'
# app types: 'app', 'apps', 'deployed'
# draft types: 'draft', 'drafts'
.
p.export()                                      # Exports all pipelines in all namespaces
p.export(namespace='NS1')                       # Exports all pipelines in namespace `NS1`
p.export(ns='NS1', type='all')                  # Exports all pipelines in namespace `NS1` -- same as above
p.export(n='default', p='app')                  # Exports deployed pipelines for namespace `default`
p.export(ns='default', type='draft')            # Exports draft pipelines for namespace `default`
p.export(ns='NS1', type='all', o='.')           # Exports all pipelines in namespace `NS1`, and save them to current directory
p.export(ns='NS1', type='deployed', o='/tmp')   # Exports draft pipelines in namespace `NS1` and save them to /tmp