A haskell implementation of the Kademlia distributed hashtable, an efficient way to store and lookup values distributed over a P2P network.

The implementation is based on the paper by Petar Maymounkov and David Mazières:br Kademlia: A Peer-to-peer Information System Based on the XOR Metric: (http://pdos.csail.mit.edu/~petar/papers/maymounkov-kademlia-lncs.pdf)

This library aims to be very simple and pleasant to use, with the downside of deciding some of the implementation details, like timeout intervals and k-bucket size, for the user.

How to use it

To get started with this library, first import it. The import has to be qualified, as the module uses the same function names as some other modules.

import qualified Network.Kademlia as K

Next, you need to decide on the types you want to use as the values to be stored in the DHT and the keys to acces them by. As soon as you've decided on them, you have to make them instances of the Serialize typeclass, so they can be sent over the network.

import qualified Data.ByteString as B
import qualified Data.ByteString.Char8 as C
import Control.Arrow (first)

-- The type this example will use as value
type Person = data {
                age :: Int
              , name :: String
              }
              deriving (Show)

instance K.Serialize Person where
   toBS = C.pack . show
   fromBS bs =
       case (reads :: ReadS Person) . C.unpack $ bs of
           [] -> Left "Failed to parse Person."
           (result, rest):_ -> Right (result, C.pack rest)

-- The type this example will use as key for the lookups
newtype KademliaID = KademliaID B.ByteString

instance K.Serialize KademliaID where
   toBS (KademliaID bs)
       | B.length bs >= 5 = B.take 5 bs
       | otherwise        = error "KademliaID to short!"

   fromBS bs
       | B.length bs >= 5 = Right . first KademliaID . B.splitAt 5 $ bs
       | otherwise        = Left "ByteString too short!"

As you could see in the example above, for the algorithm to work, you have to make sure the serialized keys are of a fixed length. There is no such constraint for the values.

Now you're ready to dive in and use the DHT:

main = do
   -- Create the first instance, which will serve as the first node of the
   -- network
   firstInstance <- K.create 12345 . KademliaID . C.pack $ "hello"

   -- Create a Node representing the first instance
   let firstNode = Node (Peer "localhost" 12345) . KademliaID . C.pack $ "hello"

   -- Create the second instance and make it join the network
   secondInstance <- K.create 12346 . KademliaID . C.pack $ "uAleu"
   joinResult <- K.joinNetwork secondInstance firstNode

   -- Make sure the joining was successful
   case joinResult of
        JoinSuccess -> do
            -- Store an example value in the network
            let exampleValue = Person 25 "Alan Turing"
            K.store secondInstance (KademliaID . C.pack $ "raxqT") exampleValue

            -- Look up the value and it's source
            (value, source) <- K.lookup firstInstance . KademliaID . C.pack $ "raxqT"
            print value

        _ -> return ()

   -- Close the instances
   K.close firstInstance
   K.close secondInstance

As promised, the usage of the actual DHT is rather easy. There are a few things to note, though:

• To join an existing network, you need to know the hostname, listening port and id of a node that is already part of that network
• When you don't need access to the DHT anymore, make sure to close the instances. This closes opened sockets and kills the threads running in the background

Another thing to note is, that you are responsible for assigning ids to nodes and keys to values, as well as making sure these are unique. The Kademlia paper doesn't propose any measures for this and, as this library is just a implementation of the system proposed in it, this library doesn't implement anything to handle this.