Quick Start

Every Fluence reference node comes with a set of builtin services which are accessible to Aqua programs. Let's use those readily available services to get the timestamp of a few of our peer-to-peer neighborhood nodes with Aqua.

{% tabs %} {% tab title="Peer Timestamps With Aqua" %}

-- timestamp_getter.aqua
-- bring the builtin services into scope
import "builtin.aqua"

-- create and identity service to join our results
service Op2("op"):
    identity(s: u64)
     array(a: string, b: u64) -> string

-- function to get ten timestamps from our Kademlia
-- neighborhood peers and return as an array of u64 timestamps
-- the function arguement node is our peer id
func ts_getter(node: string) -> []u64:
  -- create a streaming variable
  res: *[]u64
  -- execute on the pecified peer
  on node:
    -- get the base58 representation of the peer id
    k <- Op.string_to_b58(node)
    -- find all (default 20) neighborhood peers from k
    nodes <- Kademlia.neighborhood(k, false)
    -- for each peer in our neighborhood and in parallel
    for n <- nodes par:
      on n:
        -- try and get the peer's timestamp
        try:
          res <- Peer.timestamp_ms()
    -- flatten nine of our joined results
    Op2.identity(res!9)
  -- return an array of ten timestamps
  <- res

{% endtab %}

{% tab title="Compiling Aqua Script" %}

aqua-cli -i aqua-scripts -o air-scripts -a

{% endtab %}

{% tab title="Result" %}

fldist run_air -p air-scripts/timestamp_getter.ts_getter.air  -d '{"node":"12D3KooWHLxVhUQyAuZe6AHMB29P7wkvTNMn7eDMcsqimJYLKREf"}'  --generated
[
  [
    1624928596292,
    1624928596291,
    1624928596291,
    1624928596299,
    1624928596295,
    1624928596286,
    1624928596295,
    1624928596284,
    1624928596293,
    1624928596289
  ]
]

{% endtab %} {% endtabs %}

In the Aqua code, we essentially create a workflow originating from our client peer to enumerate our neighbor peers from the Kademlia neighborhood based on our reference node specification, call on the builtin timestamp service on each peer and in parallel, join after we collect ten timestamps and return our u64 array of timestamps back to the client peer.

Once we have our file, lets copy it to a directory called aqua-scripts and create an empty directory. air-scripts. Now we can compile our aqua script with the aqua-cli tool and find our AIR file in the air-scripts directory:

{% tabs %} {% tab title="Compile" %}

aqua-cli -i aqua-scripts -o air-scripts -a

{% endtab %}

{% tab title="Result" %}

# in the air-script dir you should have the following file
timestamp_getter.ts_getter.air

{% endtab %} {% endtabs %}

Once we have our AIR file we can either use Typescript or command line client. Let's use the command line client flidst:

{% tabs %} {% tab title="Run Air scripts" %}

# execute the AIR script from our compile phase with a peer id
fldist run_air -p air-scripts/timestamp_getter.ts_getter.air  -d '{"node":"12D3KooWHLxVhUQyAuZe6AHMB29P7wkvTNMn7eDMcsqimJYLKREf"}'  --generated

{% endtab %}

{% tab title="Output" %}

# here we go: ten timestamps in micro seconds obtained in parallel
[
  [
    1624928596292,
    1624928596291,
    1624928596291,
    1624928596299,
    1624928596295,
    1624928596286,
    1624928596295,
    1624928596284,
    1624928596293,
    1624928596289
  ]
]

{% endtab %} {% endtabs %}

And that's it. We now have ten reasonable timestamps from our Kademlia neighbors. See the ts-oracle example for the corresponding Aqua and Air files, respectively.

3.6 KiB Raw Blame History

Quick Start

3.6 KiB

Raw Blame History