Skip to content

Latest commit

 

History

History
475 lines (397 loc) · 14.9 KB

IMPLEMENTATION.md

File metadata and controls

475 lines (397 loc) · 14.9 KB

Implementation guidelines

How to write your own client. The Java Lavalink-Client will serve as an example implementation. The Java client has support for JDA, but can also be adapted to work with other JVM libraries.

Requirements

  • You must be able to send messages via a shard's mainWS connection.
  • You must be able to intercept voice server updates from mainWS on your shard connection.

Significant changes v2.0 -> v3.0

  • The response of /loadtracks has been completely changed (again since the initial v3.0 pre-release).
  • Lavalink v3.0 now reports its version as a handshake response header. Lavalink-Major-Version has a value of 3 for v3.0 only. It's missing for any older version.

Significant changes v1.3 -> v2.0

With the release of v2.0 many unnecessary ops were removed:

  • connect
  • disconnect
  • validationRes
  • isConnectedRes
  • validationReq
  • isConnectedReq
  • sendWS

With Lavalink 1.x the server had the responsibility of handling Discord VOICE_SERVER_UPDATEs as well as its own internal ratelimiting. This remote handling makes things unnecessarily complicated and adds a lot og points where things could go wrong. One problem we noticed is that since JDAA is unaware of ratelimits on the bot's gateway connection, it would keep adding to the ratelimit queue to the gateway. With this update this is now the responsibility of the Lavalink client or the Discord client.

A voice connection is now initiated by forwarding a voiceUpdate (VOICE_SERVER_UPDATE) to the server. When you want to disconnect or move to a different voice channel you must send Discord a new VOICE_STATE_UPDATE. If you want to move your connection to a new Lavalink server you can simply send the VOICE_SERVER_UPDATE to the new node, and the other node will be disconnected by Discord.

Depending on your Discord library, it may be possible to take advantage of the library's OP 4 handling. For instance, the JDA client takes advantage of JDA's websocket write thread to send OP 4s for connects, disconnects and reconnects.

Protocol

Opening a connection

When opening a websocket connection, you must supply 3 required headers:

Authorization: Password matching the server config
Num-Shards: Total number of shards your bot is operating on
User-Id: The user id of the bot you are playing music with

Outgoing messages

Provide an intercepted voice server update. This causes the server to connect to the voice channel.

{
    "op": "voiceUpdate",
    "guildId": "...",
    "sessionId": "...",
    "event": "..."
}

Cause the player to play a track.

startTime is an optional setting that determines the number of milliseconds to offset the track by. Defaults to 0.

endTime is an optional setting that determines at the number of milliseconds at which point the track should stop playing. Helpful if you only want to play a snippet of a bigger track. By default the track plays until it's end as per the encoded data.

If noReplace is set to true, this operation will be ignored if a track is already playing or paused.

{
    "op": "play",
    "guildId": "...",
    "track": "...",
    "startTime": "60000",
    "endTime": "120000",
    "noReplace": false
}

Cause the player to stop

{
    "op": "stop",
    "guildId": "..."
}

Set player pause

{
    "op": "pause",
    "guildId": "...",
    "pause": true
}

Make the player seek to a position of the track. Position is in millis

{
    "op": "seek",
    "guildId": "...",
    "position": 60000
}

Set player volume. Volume may range from 0 to 1000. 100 is default.

{
    "op": "volume",
    "guildId": "...",
    "volume": 125
}

Using the player equalizer

{
    "op": "equalizer",
    "guildId": "...",
    "bands": [
        {
            "band": 0,
            "gain": 0.2
        }
    ]
}

There are 15 bands (0-14) that can be changed. gain is the multiplier for the given band. The default value is 0. Valid values range from -0.25 to 1.0, where -0.25 means the given band is completely muted, and 0.25 means it is doubled. Modifying the gain could also change the volume of the output.

Tell the server to potentially disconnect from the voice server and potentially remove the player with all its data. This is useful if you want to move to a new node for a voice connection. Calling this op does not affect voice state, and you can send the same VOICE_SERVER_UPDATE to a new node.

{
    "op": "destroy",
    "guildId": "..."
}

Incoming messages

See LavalinkSocket.java for client implementation

Position information about a player. Includes unix timestamp.

{
    "op": "playerUpdate",
    "guildId": "...",
    "state": {
        "time": 1500467109,
        "position": 60000
    }
}

A collection of stats sent every minute.

{
    "op": "stats",
    ...
}

Example implementation of stats:

players = json.getInt("players");
playingPlayers = json.getInt("playingPlayers");
uptime = json.getLong("uptime");

memFree = json.getJSONObject("memory").getInt("free");
memUsed = json.getJSONObject("memory").getInt("used");
memAllocated = json.getJSONObject("memory").getInt("allocated");
memReservable = json.getJSONObject("memory").getInt("reservable");

cpuCores = json.getJSONObject("cpu").getInt("cores");
systemLoad = json.getJSONObject("cpu").getDouble("systemLoad");
lavalinkLoad = json.getJSONObject("cpu").getDouble("lavalinkLoad");

JSONObject frames = json.optJSONObject("frameStats");

if (frames != null) {
    avgFramesSentPerMinute = frames.getInt("sent");
    avgFramesNulledPerMinute = frames.getInt("nulled");
    avgFramesDeficitPerMinute = frames.getInt("deficit");
}

Server emitted an event. See the client implementation below.

{
    "op": "event",
    "type": "..."
}
/**
 * Implementation details:
 * The only events extending {@link lavalink.client.player.event.PlayerEvent} produced by the remote server are these:
 * 1. TrackStartEvent
 * 2. TrackEndEvent
 * 3. TrackExceptionEvent
 * 4. TrackStuckEvent
 * <p>
 * The remaining lavaplayer events are caused by client actions, and are therefore not forwarded via WS.
 */
private void handleEvent(JSONObject json) throws IOException {
    LavalinkPlayer player = (LavalinkPlayer) lavalink.getPlayer(json.getString("guildId"));
    PlayerEvent event = null;

    switch (json.getString("type")) {
        case "TrackStartEvent":
                event = new TrackStartEvent(player,
                        LavalinkUtil.toAudioTrack(json.getString("track"))
                );
                break;
        case "TrackEndEvent":
            event = new TrackEndEvent(player,
                    LavalinkUtil.toAudioTrack(json.getString("track")),
                    AudioTrackEndReason.valueOf(json.getString("reason"))
            );
            break;
        case "TrackExceptionEvent":
            event = new TrackExceptionEvent(player,
                    LavalinkUtil.toAudioTrack(json.getString("track")),
                    new RemoteTrackException(json.getString("error"))
            );
            break;
        case "TrackStuckEvent":
            event = new TrackStuckEvent(player,
                    LavalinkUtil.toAudioTrack(json.getString("track")),
                    json.getLong("thresholdMs")
            );
            break;
        default:
            log.warn("Unexpected event type: " + json.getString("type"));
            break;
    }

    if (event != null) player.emitEvent(event);
}

See also: AudioTrackEndReason.java

Additionally there is also the WebSocketClosedEvent, which signals when an audio web socket (to Discord) is closed. This can happen for various reasons (normal and abnormal), e.g when using an expired voice server update. 4xxx codes are usually bad. See the Discord docs.

{
    "op": "event",
    "type": "WebSocketClosedEvent",
    "guildId": "...",
    "code": 4006,
    "reason": "Your session is no longer valid.",
    "byRemote": true
}

Track Loading API

The REST api is used to resolve audio tracks for use with the play op.

GET /loadtracks?identifier=dQw4w9WgXcQ HTTP/1.1
Host: localhost:8080
Authorization: youshallnotpass

Response:

{
  "loadType": "TRACK_LOADED",
  "playlistInfo": {},
  "tracks": [
    {
      "track": "QAAAjQIAJVJpY2sgQXN0bGV5IC0gTmV2ZXIgR29ubmEgR2l2ZSBZb3UgVXAADlJpY2tBc3RsZXlWRVZPAAAAAAADPCAAC2RRdzR3OVdnWGNRAAEAK2h0dHBzOi8vd3d3LnlvdXR1YmUuY29tL3dhdGNoP3Y9ZFF3NHc5V2dYY1EAB3lvdXR1YmUAAAAAAAAAAA==",
      "info": {
        "identifier": "dQw4w9WgXcQ",
        "isSeekable": true,
        "author": "RickAstleyVEVO",
        "length": 212000,
        "isStream": false,
        "position": 0,
        "title": "Rick Astley - Never Gonna Give You Up",
        "uri": "https://www.youtube.com/watch?v=dQw4w9WgXcQ"
      }
    }
  ]
}

If the identifier leads to a playlist, playlistInfo will contain two properties, name and selectedTrack

{
  "loadType": "PLAYLIST_LOADED",
  "playlistInfo": {
    "name": "Example YouTube Playlist",
    "selectedTrack": 3
  },
  "tracks": [
    ...
  ]
}

Additionally, in every /loadtracks response, a loadType property is returned which can be used to judge the response from Lavalink properly. It can be one of the following:

  • TRACK_LOADED - Returned when a single track is loaded.
  • PLAYLIST_LOADED - Returned when a playlist is loaded.
  • SEARCH_RESULT - Returned when a search result is made (i.e ytsearch: some song).
  • NO_MATCHES - Returned if no matches/sources could be found for a given identifier.
  • LOAD_FAILED - Returned if Lavaplayer failed to load something for some reason.

If the loadType is LOAD_FAILED, the response will contain an exception object with message and severity properties. message is a string detailing why the track failed to load, and is okay to display to end-users. Severity represents how common the error is. A severity level of COMMON indicates that the error is non-fatal and that the issue is not from Lavalink itself.

{
  "loadType": "LOAD_FAILED",
  "playlistInfo": {},
  "tracks": [],
  "exception": {
    "message": "The uploader has not made this video available in your country.",
    "severity": "COMMON"
  }
}

RoutePlanner API

Additionally there are a few REST endpoints for the ip rotation extension

RoutePlanner Status

GET /routeplanner/status
Host: localhost:8080
Authorization: youshallnotpass

Response:

{
    "class": "RotatingNanoIpRoutePlanner",
    "details": {
        "ipBlock": {
            "type": "Inet6Address",
            "size": "1208925819614629174706176"
        },
        "failingAddresses": [
            {
                "address": "/1.0.0.0",
                "failingTimestamp": 1573520707545,
                "failingTime": "Mon Nov 11 20:05:07 EST 2019"
            }
        ],
        "blockIndex": "0",
        "currentAddressIndex": "36792023813"
    }
}

The response is different based on each route planner. Fields which are always present are: class, details.ipBlock and details.failingAddresses. If no route planner is set, both class and details will be null, and the other endpoints will result in status 500.

The following classes have additional detail fields:

RotatingIpRoutePlanner

details.rotateIndex String containing the number of rotations which happened since the restart of Lavalink

details.ipIndex String containing the current offset in the block

details.currentAddress The currently used ip address

NanoIpRoutePlanner

details.currentAddressIndex long representing the current offset in the ip block

RotatingNanoIpRoutePlanner

details.blockIndex String containing the the information in which /64 block ips are chosen. This number increases on each ban.

details.currentAddressIndex long representing the current offset in the ip block

Unmark a failed address

POST /routeplanner/free/address
Host: localhost:8080
Authorization: youshallnotpass

Request Body:

{
    "address": "1.0.0.1"
}

Response:

204 - No Content

Unmark all failed address

POST /routeplanner/free/all
Host: localhost:8080
Authorization: youshallnotpass

Response:

204 - No Content


All REST responses from Lavalink include a Lavalink-Api-Version header.

Resuming Lavalink sessions

What happens after your client disconnects is dependent on whether or not the session has been configured for resuming.

  • If resuming is disabled all voice connections are closed immediately.
  • If resuming is enabled all music will continue playing. You will then be able to resume your session, allowing you to control the players again.

To enable resuming, you must send a configureResuming message.

  • key is the string you will need to send when resuming the session. Set to null to disable resuming altogether. Defaults to null.
  • timeout is the number of seconds after disconnecting before the session is closed anyways. This is useful for avoiding accidental leaks. Defaults to 60 (seconds).
{
    "op": "configureResuming",
    "key": "myResumeKey",
    "timeout": 60
}

To resume a session, specify the resume key in your WebSocket handshake request headers:

Resume-Key: The resume key of the session you want to resume.

You can tell if your session was resumed by looking at the handshake response header Session-Resumed which is either true or false:

Session-Resumed: true

When a session is paused, any events that would normally have been sent is queued up. When the session is resumed, this queue is then emptied and the events are then replayed.

Special notes

  • When your shard's main WS connection dies, so does all your lavalink audio connections.

    • This also includes resumes
  • If Lavalink-Server suddenly dies (think SIGKILL) the client will have to terminate any audio connections by sending this event:

{"op":4,"d":{"self_deaf":false,"guild_id":"GUILD_ID_HERE","channel_id":null,"self_mute":false}}

Common pitfalls

Admidtedly Lavalink isn't inherently the most intuitive thing ever, and people tend to run into the same mistakes over again. Please double check the following if you run into problems developing your client and you can't connect to a voice channel or play audio:

  1. Check that you are forwarding sendWS events to Discord.
  2. Check that you are intercepting VOICE_SERVER_UPDATEs to Lavalink. Do not edit the event object from Discord.
  3. Check that you aren't expecting to hear audio when you have forgotten to queue something up OR forgotten to join a voice channel.
  4. Check that you are not trying to create a voice connection with your Discord library.
  5. When in doubt, check the debug logfile at /logs/debug.log.