Building Achlys

Cloning

Clone the Achlys repository using the following command :

$ git clone https://github.com/achlysproject/achlys.git && cd achlys && make

This should compile the framework and allow you to test it in a terminal.

Testing

Retrieve your current hostname using the following command :

$ hostname -s

And associate the result to your current IP address in the /etc/hosts file :

$ sudo nano /etc/hosts

And add a line with the pair <current IP address> <current hostname>, for example :

192.168.1.2     MyHostname ## This line has been added
127.0.0.1       localhost
255.255.255.255 broadcasthost
::1             localhost
fe80::1%lo0     localhost

NOTE : Make sure to declare it at the top of the file, since the first match will be used if any other addresses that are not currently available are declared, the Partisan communication layer will not be able to operate using the given hostname.

Launching a test shell

Configuration

If your etc/hosts file is set to resolve your local hostname to its address, that address can be directly provided to Partisan in config/test.config so that booting into Achlys will immediately allow your application to have a working communication component.

This line should be edited such that peer_ip matches your local IP address. In order to run multiple nodes on the same address, the peer_port parameter can be modified as well so that multiple instances of Partisan can run on the same host.

Running a test shell can be done with the following command from the achlys root directory :

$ make shell

In the Erlang shell, run a simulated collection of sensor data with using the Pmod_NAV module :

(achlys@Hostname)1> achlys:venom(achlys_pmod_nav_worker).

This will store the mean aggregates of temperature and pressure raw measurements, in this case the simulated values. They will be stored in CRDTs with the following Lasp identifiers :

{<<"temperature">>, state_awset}
{<<"pressure">>, state_awset}

A simple function can be created and passed to the task model in order to read the results that are currently available :

> F = fun() -> S = lasp:query({<<"temperature">>, state_awset}), sets:to_list(S) end.
> Task = achlys:declare(task1, all, permanent, F).
> achlys:bite(Task).

This will add the function in the set of distributed tasks. It can be retrieved using :

MyTask = hd(achlys:get_all_tasks()).
#{function := MyFunction} = MyTask.

Then executed using :

MyFuntion().

This should display the list of all the temperature average records that have been computed based on raw measurements and propagated once the configured aggregation trigger has been reached.