In current approach we require the following extra python libraries (some other packages such as configparser often comes with a python distribution)

• bottle.py,
• mimeparse,
• SubnetTree

Also since the packages in backends and plugins are required using the example configuration file, the path must be added to the python path before running the server. It can be done by putting the following line in files such as .bashrc:

export PYTHONPATH=$PATH_TO_ALTO_SERVER:$PATH_TO_BACKENDS:$PATH_TO_PLUGINS:$PYTHONPATH

See architecture.

Each instance MAY have multiple roles. The only one required is the HTTP service provider which is named palto.Backend currently.

Only if the backend instance wants to be included in the IRD provided by palto server, the ird section in the corresponding configuration file CAN and MUST be provided. In that circumstance, the instance has to implement the BasicIRDResource interface. The mountpoint option in ird section specifies where the resource wants to register itself.

We will be using bottle to provide RESTful service.

There are currently two backend types to be supported:

• Static file backend: reading data from a static file
• Open Daylight backend: reading data from an ODL controller

To write a customized backend, the module should contain the following method:

create_instance(resource-id, backend-config, environment)

The global configuration is now stored as environment['config'].

python -m palto.server -c examples/palto.conf

BasicIRD and some standard handlers are tested by:

python -m palto.rfc7285.irdhandler

A progress to enhance the management ability of palto is on-going. A new PaltoServer and PaltoManager are implemented. Trying them out with:

python -m palto.paltoserver

python -m palto.paltomanager

There are currently 4 basic operations on PaltoManager:

• add-backend/remove-backend: add/remove a backend from the managed PaltoServer by making a POST request to /admin/<operation>/<name>
• install/uninstall: install/uninstall a bottle plugin from either the PaltoManager or the PaltoServer, depending on the target parameter, by making a POST request to /admin/<operation>/<target>/<name>

The content of the POST request would be treated as a configuration file. All the operations would require a provider option in the basic section. The provider module must come with a create_instance method which takes the name, config(parsed from the request) and global_config(initialized when the manager is launched and maintained by the manager).

And you can use the following commands to test them out:

# for paltoserver

## should get 501
curl -D - -X GET http://localhost:3400/test

## should get 200
curl -D - -X GET http://localhost:3400/get_baseurl

# for paltomanager

## should get 404
curl -D -X GET http://localhost:3400/alto/test_sf_networkmap

## should get 200
curl -D - -X POST -u test:test \
      --data-binary @examples/input/test_staticfile.conf \
      http://localhost:3400/admin/add-backend/test_sf_networkmap

## should get 200
curl -D - -X GET http://localhost:3400/alto/test_sf_networkmap

## should get 200
curl -D - -X POST -u test:test \
      http://localhost:3400/admin/remove-backend/test_sf_networkmap

## should get 200
curl -D - -X POST -u test:test \
      --data-binary @examples/input/test_plugin.conf \
      http://localhost:3400/admin/install/paltoserver/test_plugin

## should get 200 and see some output in the server console
curl -D - -X GET http://localhost:3400/alto/test_sf_networkmap

## should get 200
curl -D - -X POST -u test:test \
      http://localhost:3400/admin/uninstall/paltoserver/test_plugin

## Install AutoIRDPlugin
## should get 200
curl -D - -X POST -u test:test \
      --data-binary @examples/input/test_autoird.conf \
      http://localhost:3400/admin/install/paltomanager/autoird

## should get 200 and see the generated root IRD
curl -D - -X GET http://localhost:3400/alto/

## should get 200 and see no output in the server console
curl -D - -X GET http://localhost:3400/alto/test_sf_networkmap

## should get 404 again
curl -D -X GET http://localhost:3400/alto/test_sf_networkmap

The support for network map is implemented. The example data file can be found in examples/static_file/networkmap.json.

Run the following command in one terminal to start the server:

python -m palto.server -c examples/palto.conf

Then run the following command in another terminal to see the result:

curl -D - -X GET http://localhost:3400/test_sf_networkmap
curl -D - -X GET http://localhost:3400/test_sf_costmap

You can also run the following command to see the generated ird:

curl -D - -X GET http://localhost:3400/

A simple implementation for several services.

See ECSLite

A simple IRD generation plugin for palto.

See AutoIRDPlugin for details.

• Improve the code structure for server/backend/frontend (on-going)
• Management interface for palto (on-going)
• Dependency management and tracing (especially for ODL-backend)
• Tag support