Add config for the shard start wait time (#3031)

Co-authored-by: Gnome! <[email protected]>

src/gateway/client/mod.rs

@@ -33,6 +33,7 @@ use std::ops::Range;
 use std::sync::Arc;
 #[cfg(feature = "framework")]
 use std::sync::OnceLock;
+use std::time::Duration;
 
 use futures::channel::mpsc::UnboundedReceiver as Receiver;
 use futures::future::BoxFuture;
@@ -49,7 +50,14 @@ use crate::cache::Settings as CacheSettings;
 use crate::framework::Framework;
 #[cfg(feature = "voice")]
 use crate::gateway::VoiceGatewayManager;
-use crate::gateway::{ActivityData, GatewayError, PresenceData, ShardManager, ShardManagerOptions};
+use crate::gateway::{
+    ActivityData,
+    GatewayError,
+    PresenceData,
+    ShardManager,
+    ShardManagerOptions,
+    DEFAULT_WAIT_BETWEEN_SHARD_START,
+};
 use crate::http::Http;
 use crate::internal::prelude::*;
 use crate::internal::tokio::spawn_named;
@@ -73,6 +81,7 @@ pub struct ClientBuilder {
     event_handler: Option<Arc<dyn EventHandler>>,
     raw_event_handler: Option<Arc<dyn RawEventHandler>>,
     presence: PresenceData,
+    wait_time_between_shard_start: Duration,
 }
 
 impl ClientBuilder {
@@ -106,6 +115,7 @@ impl ClientBuilder {
             event_handler: None,
             raw_event_handler: None,
             presence: PresenceData::default(),
+            wait_time_between_shard_start: DEFAULT_WAIT_BETWEEN_SHARD_START,
         }
     }
 
@@ -153,6 +163,19 @@ impl ClientBuilder {
         self.framework.as_deref()
     }
 
+    /// Sets the time to wait between starting shards.
+    ///
+    /// This should only be used when using a gateway proxy, such as [Sandwich] or [Twilight Gateway
+    /// Proxy], as otherwise this will lead to gateway disconnects if the shard start rate limit is
+    /// not respected.
+    ///
+    /// [Sandwich]:  https://github.com/WelcomerTeam/Sandwich-Daemon
+    /// [Twilight Gateway Proxy]: https://github.com/Gelbpunkt/gateway-proxy
+    pub fn wait_time_between_shard_start(mut self, wait_time: Duration) -> Self {
+        self.wait_time_between_shard_start = wait_time;
+        self
+    }
+
     /// Sets the voice gateway handler to be used. It will receive voice events sent over the
     /// gateway and then consider - based on its settings - whether to dispatch a command.
     #[cfg(feature = "voice")]
@@ -318,6 +341,7 @@ impl IntoFuture for ClientBuilder {
                 intents,
                 presence: Some(presence),
                 max_concurrency,
+                wait_time_between_shard_start: self.wait_time_between_shard_start,
             });
 
             let client = Client {

src/gateway/sharding/mod.rs

@@ -50,7 +50,11 @@ use tokio_tungstenite::tungstenite::protocol::frame::CloseFrame;
 use tracing::{debug, error, info, trace, warn};
 use url::Url;
 
-pub use self::shard_manager::{ShardManager, ShardManagerOptions};
+pub use self::shard_manager::{
+    ShardManager,
+    ShardManagerOptions,
+    DEFAULT_WAIT_BETWEEN_SHARD_START,
+};
 pub use self::shard_messenger::ShardMessenger;
 pub use self::shard_queuer::{ShardQueue, ShardQueuer, ShardQueuerMessage};
 pub use self::shard_runner::{ShardRunner, ShardRunnerMessage, ShardRunnerOptions};

src/gateway/sharding/shard_manager.rs

@@ -25,6 +25,9 @@ use crate::internal::prelude::*;
 use crate::internal::tokio::spawn_named;
 use crate::model::gateway::GatewayIntents;
 
+/// The default time to wait between starting each shard or set of shards.
+pub const DEFAULT_WAIT_BETWEEN_SHARD_START: Duration = Duration::from_secs(5);
+
 /// A manager for handling the status of shards by starting them, restarting them, and stopping
 /// them when required.
 ///
@@ -50,7 +53,7 @@ use crate::model::gateway::GatewayIntents;
 /// use std::sync::{Arc, OnceLock};
 ///
 /// use serenity::gateway::client::EventHandler;
-/// use serenity::gateway::{ShardManager, ShardManagerOptions};
+/// use serenity::gateway::{ShardManager, ShardManagerOptions, DEFAULT_WAIT_BETWEEN_SHARD_START};
 /// use serenity::http::Http;
 /// use serenity::model::gateway::GatewayIntents;
 /// use serenity::prelude::*;
@@ -84,6 +87,7 @@ use crate::model::gateway::GatewayIntents;
 ///     intents: GatewayIntents::non_privileged(),
 ///     presence: None,
 ///     max_concurrency,
+///     wait_time_between_shard_start: DEFAULT_WAIT_BETWEEN_SHARD_START,
 /// });
 /// # Ok(())
 /// # }
@@ -146,6 +150,7 @@ impl ShardManager {
             http: opt.http,
             intents: opt.intents,
             presence: opt.presence,
+            wait_time_between_shard_start: opt.wait_time_between_shard_start,
         };
 
         spawn_named("shard_queuer::run", async move {
@@ -372,4 +377,6 @@ pub struct ShardManagerOptions {
     pub intents: GatewayIntents,
     pub presence: Option<PresenceData>,
     pub max_concurrency: NonZeroU16,
+    /// Number of seconds to wait between starting each shard/set of shards start
+    pub wait_time_between_shard_start: Duration,
 }

src/gateway/sharding/shard_queuer.rs

@@ -31,8 +31,6 @@ use crate::internal::prelude::*;
 use crate::internal::tokio::spawn_named;
 use crate::model::gateway::{GatewayIntents, ShardInfo};
 
-const WAIT_BETWEEN_BOOTS_IN_SECONDS: u64 = 5;
-
 /// The shard queuer is a simple loop that runs indefinitely to manage the startup of shards.
 ///
 /// A shard queuer instance _should_ be run in its own thread, due to the blocking nature of the
@@ -68,6 +66,8 @@ pub struct ShardQueuer {
     pub ws_url: Arc<str>,
     /// The total amount of shards to start.
     pub shard_total: NonZeroU16,
+    /// Number of seconds to wait between each start
+    pub wait_time_between_shard_start: Duration,
     #[cfg(feature = "cache")]
     pub cache: Arc<Cache>,
     pub http: Arc<Http>,
@@ -94,14 +94,14 @@ impl ShardQueuer {
     /// **Note**: This should be run in its own thread due to the blocking nature of the loop.
     #[cfg_attr(feature = "tracing_instrument", instrument(skip(self)))]
     pub async fn run(&mut self) {
-        // We read from the Rx channel in a loop, and use a timeout of 5 seconds so that we don't
+        // We read from the Rx channel in a loop, and use a timeout of
+        // {self.WAIT_TIME_BETWEEN_SHARD_START} (5 seconds normally) seconds so that we don't
         // hang forever. When we receive a command to start a shard, we append it to our queue. The
         // queue is popped in batches of shards, which are started in parallel. A batch is fired
-        // every 5 seconds at minimum in order to avoid being ratelimited.
-        const TIMEOUT: Duration = Duration::from_secs(WAIT_BETWEEN_BOOTS_IN_SECONDS);
+        // every WAIT_TIME_BETWEEN_SHARD_START at minimum in order to avoid being ratelimited.
 
         loop {
-            if let Ok(msg) = timeout(TIMEOUT, self.rx.next()).await {
+            if let Ok(msg) = timeout(self.wait_time_between_shard_start, self.rx.next()).await {
                 match msg {
                     Some(ShardQueuerMessage::SetShardTotal(shard_total)) => {
                         self.shard_total = shard_total;
@@ -157,14 +157,13 @@ impl ShardQueuer {
         let Some(instant) = self.last_start else { return };
 
         // We must wait 5 seconds between IDENTIFYs to avoid session invalidations.
-        let duration = Duration::from_secs(WAIT_BETWEEN_BOOTS_IN_SECONDS);
         let elapsed = instant.elapsed();
 
-        if elapsed >= duration {
+        if elapsed >= self.wait_time_between_shard_start {
             return;
         }
 
-        let to_sleep = duration - elapsed;
+        let to_sleep = self.wait_time_between_shard_start - elapsed;
 
         sleep(to_sleep).await;
     }