lfdversluis/fddg

Fault-tolerant Distributed Dragon Game - distributed computing systems lab

fddg's Issues

Implement what to do when a server or client is down (heartbeat fails)

Think out the model for connecting/disconnecting clients

Since one of the requirements of the game is to implement a realistic pattern of connecting/disconnection actions from clients and crashing of servers, we should come up with a model for this. The assignment has two pointers to existing literature:

http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6404027
http://www.pds.ewi.tudelft.nl/~iosup/game-trace-archive12netgames.pdf

Introduction

Implement the Field class

Change code so that only updates are sent, rather than the whole field

Write about the client simulation in the report

We should dedicate a section in the report about the client simulation. I think this is a good opportunity to actually make some references.

Write about Experimental setup in the report

Remove unacknowledged requests

Currently, requests from a server of which not all acks are received are never removed from the buffer. We should probably implement this using a timer for each request (of for instance 5s). When all acks are received, the timer should be stopped. When it expires, the request should be removed

Conclusion

Implement Dragon class

Implement the Player class

Players should be able to die.

Let clients select a random server

Now, all clients simple try to connect to the first server in the setup/servers.txt file. However, they should randomly pick one of them and try to connect to it. When this fails, it should randomly select another server.

Implement the protocol for communication between the servers

Write Background on Application in the report

Write about experimental results and improve discussion section based on the results.

Create a diagram of the system architecture with draw.io

Front page

Comments on current code

Here are some comments I have on the current codebase:

• The package names (nl.tud) are not really unique. Something like nl.tud.dcs.fddg will be much better. (fixed in #29)
• The code is currently not documented that well. I think that at least all interface methods (and preferably all other public methods as well) need a JavaDoc comment to inform each other about what the method does and why it is there (so it also documents the protocol). Now, you cannot see what some methods actually do without examining the code. (Fixed in #32)
• With the current Main class you can only start all servers and clients on the same machine. However, we need to be able to distribute them across multiple machines to actually make it a distributed system. I suggest we do this by having a main method that only starts one server on the local machine. Then we can actually start each server on another machine. The same holds for the client processes. We can then also use scripts to for instance start x clients on one machine. (fixed in #31)
• In the current Main class the IDs of the clients are randomly generated. This is not really correct as multiple clients can now have the same ID. Instead, the clients should receive their ID from the server (and not pick one themselves). (fixed in #39)
• I think the GUI of a server should be optional so that you can also start one without a GUI. Currently, a new GUI is created for every server.(fixed in #31)
• Why do we wait 5 seconds (and 1 second in the loop) in ServerProcess.run()? This is an example of where a comment should clarify this ;) (fixed in 2082fd7)
• This (in ServerProcess.connect()) is not going to work when the clients run on different machines: ClientInterface ci = (ClientInterface) Naming.lookup("rmi://localhost:" + Main.SERVER_PORT + "/FDDGClient/" + playerId); (fixed in #47)
• ServerProcess.broadcastFieldToConnectedPlayers() does not meet our design as the entire field is only sent once to each client (when he connects), so it will never be broadcast. This is probably related to issue #24
• I think the XAction classes do not really represent game objects, so I would put them in another package. (fixed in #29)
• For creating a copy of the path in Field.getDirectionToNearestDragon() you can better use the copy constructor of ArrayList (so new ArrayList<>(path)) instead of: ArrayList<Integer> updatedPath = new ArrayList<Integer>(path.size()); for(int pos : path){ updatedPath.add(pos); } (fixed in #29)
• The Field.dragonRage() function should be replaced because this function only works when there is only one server. With multiple servers, the game state will become inconsistent. Perhaps only perform the rage on the players that have connected to the particular server. Then there is no need for sending messages to other servers. (fixed in #47)

Players should be able to connect dynamically to a server.

RMI exception with different hosts (>=1 non-Windows)

With two hosts of which at least one does not run Windows when you start one client, you get the following exception:

INFO: Starting client
mrt 22, 2015 4:55:38 PM nl.tud.dcs.fddg.client.ClientProcess selectServer
INFO: Client 0 trying to connect to //localhost:1099/FDDGServer/0
java.rmi.ServerException: RemoteException occurred in server thread; nested exception is:
        java.rmi.ConnectException: Connection refused to host: 127.0.1.1; nested exception is:
        java.net.ConnectException: Connection refused: connect
        at sun.rmi.server.UnicastServerRef.dispatch(UnicastServerRef.java:354)
        at sun.rmi.transport.Transport$1.run(Transport.java:200)
        at sun.rmi.transport.Transport$1.run(Transport.java:197)
        at java.security.AccessController.doPrivileged(Native Method)
        at sun.rmi.transport.Transport.serviceCall(Transport.java:196)
        at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:568)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run0(TCPTransport.java:826)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.lambda$run$240(TCPTransport.java:683)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler$$Lambda$17/1664698779.run(Unknown Source)
        at java.security.AccessController.doPrivileged(Native Method)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run(TCPTransport.java:682)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
        at java.lang.Thread.run(Thread.java:745)
        at sun.rmi.transport.StreamRemoteCall.exceptionReceivedFromServer(Unknown Source)
        at sun.rmi.transport.StreamRemoteCall.executeCall(Unknown Source)
        at sun.rmi.server.UnicastRef.invoke(Unknown Source)
        at java.rmi.server.RemoteObjectInvocationHandler.invokeRemoteMethod(Unknown Source)
        at java.rmi.server.RemoteObjectInvocationHandler.invoke(Unknown Source)
        at com.sun.proxy.$Proxy1.connect(Unknown Source)
        at nl.tud.dcs.fddg.client.ClientProcess.run(ClientProcess.java:112)
        at java.lang.Thread.run(Unknown Source)
Caused by: java.rmi.ConnectException: Connection refused to host: 127.0.1.1; nested exception is:
        java.net.ConnectException: Connection refused: connect
        at sun.rmi.transport.tcp.TCPEndpoint.newSocket(TCPEndpoint.java:619)
        at sun.rmi.transport.tcp.TCPChannel.createConnection(TCPChannel.java:216)
        at sun.rmi.transport.tcp.TCPChannel.newConnection(TCPChannel.java:202)
        at sun.rmi.server.UnicastRef.invoke(UnicastRef.java:130)
        at java.rmi.server.RemoteObjectInvocationHandler.invokeRemoteMethod(RemoteObjectInvocationHandler.java:194)
        at java.rmi.server.RemoteObjectInvocationHandler.invoke(RemoteObjectInvocationHandler.java:148)
        at com.sun.proxy.$Proxy0.performAction(Unknown Source)
        at nl.tud.dcs.fddg.server.ServerProcess.broadcastActionToServers(ServerProcess.java:363)
        at nl.tud.dcs.fddg.server.ServerProcess.connect(ServerProcess.java:234)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.lang.reflect.Method.invoke(Method.java:483)
        at sun.rmi.server.UnicastServerRef.dispatch(UnicastServerRef.java:323)
        at sun.rmi.transport.Transport$1.run(Transport.java:200)
        at sun.rmi.transport.Transport$1.run(Transport.java:197)
        at java.security.AccessController.doPrivileged(Native Method)
        at sun.rmi.transport.Transport.serviceCall(Transport.java:196)
        at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:568)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run0(TCPTransport.java:826)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.lambda$run$240(TCPTransport.java:683)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler$$Lambda$17/1664698779.run(Unknown Source)
        at java.security.AccessController.doPrivileged(Native Method)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run(TCPTransport.java:682)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
        at java.lang.Thread.run(Thread.java:745)
Caused by: java.net.ConnectException: Connection refused: connect
        at java.net.DualStackPlainSocketImpl.connect0(Native Method)
        at java.net.DualStackPlainSocketImpl.socketConnect(DualStackPlainSocketImpl.java:79)
        at java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:345)
        at java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:206)
        at java.net.AbstractPlainSocketImpl.connect(AbstractPlainSocketImpl.java:188)
        at java.net.PlainSocketImpl.connect(PlainSocketImpl.java:172)
        at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:392)
        at java.net.Socket.connect(Socket.java:589)
        at java.net.Socket.connect(Socket.java:538)
        at java.net.Socket.<init>(Socket.java:434)
        at java.net.Socket.<init>(Socket.java:211)
        at sun.rmi.transport.proxy.RMIDirectSocketFactory.createSocket(RMIDirectSocketFactory.java:40)
        at sun.rmi.transport.proxy.RMIMasterSocketFactory.createSocket(RMIMasterSocketFactory.java:148)
        at sun.rmi.transport.tcp.TCPEndpoint.newSocket(TCPEndpoint.java:613)
        ... 26 more

Discussion

Implement heartbeat mechanism between servers

Implement the protocol for communication between the client and server

Implement what to do when a server "peer" drops.

If a server pings a peer server and it fails two times, what do you do?

Check multiple attack or heal action

Following the discussion in #47, we should remove the checks for multiple heal or actions and see what happens

Write about experiments in the report

Incorrect dragonRage function

Now the dragonRage function looks as follows:

Set<Action> actionSet = new HashSet<Action>();

        for (int dragonId : dragonMap.keySet()) {
            Dragon d = dragonMap.get(dragonId);

            int dragonX = d.getxPos();
            int dragonY = d.getyPos();

            for (int i = 0; i < 4; i++) {
                int unitX = dragonX + dx[i];
                int unitY = dragonY + dy[i];

                if (isInBoard(unitX, unitY) && entities[unitY][unitX] instanceof Player && connectedPlayers.contains(entities[unitY][unitX].getUnitId())) {
                    Player p = (Player) entities[unitY][unitX];
                    p.setCurHitPoints(p.getCurHitPoints() - d.getAttackPower());

                    if (p.getCurHitPoints() <= 0) {
                        DeleteUnitAction dua = new DeleteUnitAction(p.getUnitId());
                        actionSet.add(dua);
                        removePlayer(p.getUnitId());
                    }

                    DamageAction da = new DamageAction(p.getUnitId(), d.getAttackPower());
                    actionSet.add(da);
                }
            }
        }
        return actionSet;

Dragon can attack players when the distance between them is at most 2. However, in the code above the for-loop doesn't do this correctly, as it only seems to search for players that are to right and above the dragon (and not also below or to the left). Also the range for searching these players is to big, as it can only be 2 at most (so horizontalDistance+vertDistance <= 2)

Make dragons attack players

System Design, System overview

NPE in Field.isInRange()

After a while (with 100 clients), I get the following exception in multiple client processes:

Exception in thread "Thread-43" java.lang.NullPointerException
        at nl.tud.dcs.fddg.game.Field.isInRange(Field.java:108)
        at nl.tud.dcs.fddg.server.ServerProcess.attack(ServerProcess.java:128)
        at nl.tud.dcs.fddg.server.ServerProcess.performAction(ServerProcess.java:91)
        at sun.reflect.GeneratedMethodAccessor9.invoke(Unknown Source)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)
        at java.lang.reflect.Method.invoke(Unknown Source)
        at sun.rmi.server.UnicastServerRef.dispatch(Unknown Source)
        at sun.rmi.transport.Transport$1.run(Unknown Source)
        at sun.rmi.transport.Transport$1.run(Unknown Source)
        at java.security.AccessController.doPrivileged(Native Method)
        at sun.rmi.transport.Transport.serviceCall(Unknown Source)
        at sun.rmi.transport.tcp.TCPTransport.handleMessages(Unknown Source)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run0(Unknown Source)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.lambda$run$240(Unknown Source)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler$$Lambda$17/833442319.run(Unknown Source)
        at java.security.AccessController.doPrivileged(Native Method)
        at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run(Unknown Source)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
        at java.lang.Thread.run(Unknown Source)
        at sun.rmi.transport.StreamRemoteCall.exceptionReceivedFromServer(Unknown Source)
        at sun.rmi.transport.StreamRemoteCall.executeCall(Unknown Source)
        at sun.rmi.server.UnicastRef.invoke(Unknown Source)
        at java.rmi.server.RemoteObjectInvocationHandler.invokeRemoteMethod(Unknown Source)
        at java.rmi.server.RemoteObjectInvocationHandler.invoke(Unknown Source)
        at com.sun.proxy.$Proxy1.performAction(Unknown Source)
        at nl.tud.dcs.fddg.client.ClientProcess.run(ClientProcess.java:82)
        at java.lang.Thread.run(Unknown Source)

Write an abstract for the report

Make sure to client don't get the same ID

Now, when two clients connect to two different servers, both can receive the same ID which leads to a field with two players with the same ID.

HealAction with only one client

Now, when you start only one client, it also sends requests for a HealAction to the server. However, this should not be possible with one client (as there is no other player to heal).

Exception is thrown when a client connects to one server (setup: 2 servers, 1 client)

java.rmi.NotBoundException: FDDGClient/0
    at sun.rmi.registry.RegistryImpl.lookup(RegistryImpl.java:136)
    at sun.rmi.registry.RegistryImpl_Skel.dispatch(Unknown Source)
    at sun.rmi.server.UnicastServerRef.oldDispatch(UnicastServerRef.java:409)
    at sun.rmi.server.UnicastServerRef.dispatch(UnicastServerRef.java:267)
    at sun.rmi.transport.Transport$1.run(Transport.java:177)
    at sun.rmi.transport.Transport$1.run(Transport.java:174)
    at java.security.AccessController.doPrivileged(Native Method)
    at sun.rmi.transport.Transport.serviceCall(Transport.java:173)
    at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:556)
    at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run0(TCPTransport.java:811)
    at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run(TCPTransport.java:670)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
    at java.lang.Thread.run(Thread.java:745)
    at sun.rmi.transport.StreamRemoteCall.exceptionReceivedFromServer(StreamRemoteCall.java:275)
    at sun.rmi.transport.StreamRemoteCall.executeCall(StreamRemoteCall.java:252)
    at sun.rmi.server.UnicastRef.invoke(UnicastRef.java:378)
    at sun.rmi.registry.RegistryImpl_Stub.lookup(Unknown Source)
    at java.rmi.Naming.lookup(Naming.java:101)
    at nl.tud.dcs.fddg.server.ServerProcess.reconnect(ServerProcess.java:326)
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
    at java.lang.reflect.Method.invoke(Method.java:606)
    at sun.rmi.server.UnicastServerRef.dispatch(UnicastServerRef.java:322)
    at sun.rmi.transport.Transport$1.run(Transport.java:177)
    at sun.rmi.transport.Transport$1.run(Transport.java:174)
    at java.security.AccessController.doPrivileged(Native Method)
    at sun.rmi.transport.Transport.serviceCall(Transport.java:173)
    at sun.rmi.transport.tcp.TCPTransport.handleMessages(TCPTransport.java:556)
    at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run0(TCPTransport.java:811)
    at sun.rmi.transport.tcp.TCPTransport$ConnectionHandler.run(TCPTransport.java:670)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
    at java.lang.Thread.run(Thread.java:745)

Dragons are not losing health when using 2 servers and 1 player

It could be a GUI issue maybe?

Implement the functionality when a server drops and comes back up

• Connect to peers, let them know you are back
• Receive the field
• Resume participating
• Balance load?

Additional System Features

Clients on different host than the server

Now, clients can only be run on the same host as the server they connect to. However, clients should be able to connect to servers that run on other hosts

Server stopping when one of the clients are disconnecting

The servers are stopping if one of the clients disconnect.

Write about the pros and cons of our approach regarding consistency, fault tolerance and replication

We should dedicate a short section of our report discussing the pros and cons of the decisions we made about:
-replication
-fault tolerance
-consistency

Implement the Player class

Detecting whether the game has finished

When a sever has dropped and sends a ping, make the connection again to start sending updates (and the board on initial start)

Implement the player strategy

Repository ownership

I am perfectly fine with keeping the repository on my account until you want to open source it, but if anyone else has a slot for a private repository and wants to take ownership, let me know 😉

Reconnecting to another server doesn't work correctly

In the current code, when the server a client was connected to crashes, it selects another server. However, this client is never added to the connectedPlayers map in the new server (as connect() is not called). In this way it is never attacked by dragons anymore

Broadcast end of game

When one of the servers decides to finish the game, it should broadcast this to all other servers and clients. The other servers and clients can then also finish the game (and exit the program).

Implement the realistic simulation of connecting/disconnecting players

Now we can generate files for simulation, the simulation file should be read and executed. This can for example be done in Java or Bash.

Bug bug bug!

a player can hit a dragon (or vice versa) from a distance of at most 2 squares.

We implemented 1 :(