sender_workers scaling. this time, with feeling!

scale rocksdb background jobs and subcompactions

change rocksdb default error level to info from error

delete unused num_threads function

fix warns from cargo

.forgejo/workflows/release-image.yml

@@ -49,6 +49,7 @@ jobs:
             const platforms = ['linux/amd64', 'linux/arm64']
             core.setOutput('build_matrix', JSON.stringify({
               platform: platforms,
+              target_cpu: ['base'],
               include: platforms.map(platform => { return {
                 platform,
                 slug: platform.replace('/', '-')
@@ -66,6 +67,8 @@ jobs:
     strategy:
       matrix:
         {
+          "target_cpu": ["base"],
+          "profile": ["release"],
           "include":
             [
               { "platform": "linux/amd64", "slug": "linux-amd64" },
@@ -73,6 +76,7 @@ jobs:
             ],
           "platform": ["linux/amd64", "linux/arm64"],
         }
+
     steps:
       - name: Echo strategy
         run: echo '${{ toJSON(fromJSON(needs.define-variables.outputs.build_matrix)) }}'
@@ -140,8 +144,8 @@ jobs:
         uses: actions/cache@v3
         with:
           path: |
-            cargo-target-${{ matrix.slug }}
-          key: cargo-target-${{ matrix.slug }}-${{hashFiles('**/Cargo.lock') }}-${{steps.rust-toolchain.outputs.rustc_version}}
+            cargo-target-${{ matrix.target_cpu }}-${{ matrix.slug }}-${{ matrix.profile }}
+          key: cargo-target-${{ matrix.target_cpu }}-${{ matrix.slug }}-${{ matrix.profile }}-${{hashFiles('**/Cargo.lock') }}-${{steps.rust-toolchain.outputs.rustc_version}}
       - name: Cache apt cache
         id: cache-apt
         uses: actions/cache@v3
@@ -163,9 +167,9 @@ jobs:
             {
               ".cargo/registry": "/usr/local/cargo/registry",
               ".cargo/git/db": "/usr/local/cargo/git/db",
-              "cargo-target-${{ matrix.slug }}": {
+              "cargo-target-${{ matrix.target_cpu }}-${{ matrix.slug }}-${{ matrix.profile }}": {
                 "target": "/app/target",
-                "id": "cargo-target-${{ matrix.platform }}"
+                "id": "cargo-target-${{ matrix.target_cpu }}-${{ matrix.slug }}-${{ matrix.profile }}"
               },
               "var-cache-apt-${{ matrix.slug }}": "/var/cache/apt",
               "var-lib-apt-${{ matrix.slug }}": "/var/lib/apt"

conduwuit-example.toml

@@ -990,7 +990,7 @@
 # 3 to 5 = Statistics with possible performance impact.
 # 6 = All statistics.
 #
-#rocksdb_stats_level = 1
+#rocksdb_stats_level = 3
 
 # This is a password that can be configured that will let you login to the
 # server bot account (currently `@conduit`) for emergency troubleshooting
@@ -1590,11 +1590,9 @@
 #stream_amplification = 1024
 
 # Number of sender task workers; determines sender parallelism. Default is
-# '0' which means the value is determined internally, likely matching the
-# number of tokio worker-threads or number of cores, etc. Override by
-# setting a non-zero value.
+# number of CPU cores. Override by setting a different value.
 #
-#sender_workers = 0
+#sender_workers = 4
 
 # Enables listener sockets; can be set to false to disable listening. This
 # option is intended for developer/diagnostic purposes only.

docker/Dockerfile

@@ -1,15 +1,16 @@
 ARG RUST_VERSION=1
+ARG DEBIAN_VERSION=bookworm
 
 FROM --platform=$BUILDPLATFORM docker.io/tonistiigi/xx AS xx
-FROM --platform=$BUILDPLATFORM rust:${RUST_VERSION}-slim-bookworm AS base
-FROM --platform=$BUILDPLATFORM rust:${RUST_VERSION}-slim-bookworm AS toolchain
+FROM --platform=$BUILDPLATFORM rust:${RUST_VERSION}-slim-${DEBIAN_VERSION} AS base
+FROM --platform=$BUILDPLATFORM rust:${RUST_VERSION}-slim-${DEBIAN_VERSION} AS toolchain
 
 # Prevent deletion of apt cache
 RUN rm -f /etc/apt/apt.conf.d/docker-clean
 
 # Match Rustc version as close as possible
 # rustc -vV
-ARG LLVM_VERSION=19
+ARG LLVM_VERSION=20
 # ENV RUSTUP_TOOLCHAIN=${RUST_VERSION}
 
 # Install repo tools
@@ -19,10 +20,18 @@ ARG LLVM_VERSION=19
 RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
     --mount=type=cache,target=/var/lib/apt,sharing=locked \
     apt-get update && apt-get install -y \
-    clang-${LLVM_VERSION} lld-${LLVM_VERSION} pkg-config make jq \
-    curl git \
+    pkg-config make jq \
+    curl git software-properties-common \
     file
 
+# LLVM packages
+RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
+    --mount=type=cache,target=/var/lib/apt,sharing=locked \
+    curl https://apt.llvm.org/llvm.sh > llvm.sh && \
+    chmod +x llvm.sh && \
+    ./llvm.sh ${LLVM_VERSION} && \
+    rm llvm.sh
+
 # Create symlinks for LLVM tools
 RUN <<EOF
     set -o xtrace
@@ -39,7 +48,7 @@ EOF
 
 # Developer tool versions
 # renovate: datasource=github-releases depName=cargo-bins/cargo-binstall
-ENV BINSTALL_VERSION=1.12.3
+ENV BINSTALL_VERSION=1.13.0
 # renovate: datasource=github-releases depName=psastras/sbom-rs
 ENV CARGO_SBOM_VERSION=0.9.1
 # renovate: datasource=crate depName=lddtree
@@ -140,11 +149,12 @@ ENV GIT_REMOTE_COMMIT_URL=$GIT_REMOTE_COMMIT_URL
 ENV CONDUWUIT_VERSION_EXTRA=$CONDUWUIT_VERSION_EXTRA
 ENV CONTINUWUITY_VERSION_EXTRA=$CONTINUWUITY_VERSION_EXTRA
 
+ARG RUST_PROFILE=release
 
 # Build the binary
 RUN --mount=type=cache,target=/usr/local/cargo/registry \
     --mount=type=cache,target=/usr/local/cargo/git/db \
-    --mount=type=cache,target=/app/target,id=cargo-target-${TARGETPLATFORM} \
+    --mount=type=cache,target=/app/target,id=cargo-target-${TARGET_CPU}-${TARGETPLATFORM}-${RUST_PROFILE} \
     bash <<'EOF'
     set -o allexport
     set -o xtrace
@@ -153,7 +163,7 @@ RUN --mount=type=cache,target=/usr/local/cargo/registry \
             jq -r ".target_directory"))
     mkdir /out/sbin
     PACKAGE=conduwuit
-    xx-cargo build --locked --release \
+    xx-cargo build --locked --profile ${RUST_PROFILE} \
         -p $PACKAGE;
     BINARIES=($(cargo metadata --no-deps --format-version 1 | \
         jq -r ".packages[] | select(.name == \"$PACKAGE\") | .targets[] | select( .kind | map(. == \"bin\") | any ) | .name"))

docs/deploying/arch-linux.md

@@ -1,3 +1,5 @@
 # Continuwuity for Arch Linux
 
-Continuwuity does not have any Arch Linux packages at this time.
+Continuwuity is available on the `archlinuxcn` repository and AUR, with the same package name `continuwuity`, which includes latest taggged version. The development version is available on AUR as `continuwuity-git`
+
+Simply install the `continuwuity` package. Configure the service in `/etc/conduwuit/conduwuit.toml`, then enable/start the continuwuity.service.

rust-toolchain.toml

@@ -9,7 +9,7 @@
 # If you're having trouble making the relevant changes, bug a maintainer.
 
 [toolchain]
-channel = "1.86.0"
+channel = "1.87.0"
 profile = "minimal"
 components = [
     # For rust-analyzer
@@ -19,11 +19,3 @@ components = [
     "rustfmt",
     "clippy",
 ]
-targets = [
-    #"x86_64-apple-darwin",
-    "x86_64-unknown-linux-gnu",
-    "x86_64-unknown-linux-musl",
-    "aarch64-unknown-linux-musl",
-    "aarch64-unknown-linux-gnu",
-    #"aarch64-apple-darwin",
-]

src/admin/debug/commands.rs

@@ -239,10 +239,11 @@ pub(super) async fn get_remote_pdu(
 		})
 		.await
 	{
-		| Err(e) =>
+		| Err(e) => {
 			return Err!(
 				"Remote server did not have PDU or failed sending request to remote server: {e}"
-			),
+			);
+		},
 		| Ok(response) => {
 			let json: CanonicalJsonObject =
 				serde_json::from_str(response.pdu.get()).map_err(|e| {
@@ -384,8 +385,9 @@ pub(super) async fn change_log_level(&self, filter: Option<String>, reset: bool)
 			.reload
 			.reload(&old_filter_layer, Some(handles))
 		{
-			| Err(e) =>
-				return Err!("Failed to modify and reload the global tracing log level: {e}"),
+			| Err(e) => {
+				return Err!("Failed to modify and reload the global tracing log level: {e}");
+			},
 			| Ok(()) => {
 				let value = &self.services.server.config.log;
 				let out = format!("Successfully changed log level back to config value {value}");
@@ -408,8 +410,9 @@ pub(super) async fn change_log_level(&self, filter: Option<String>, reset: bool)
 			.reload(&new_filter_layer, Some(handles))
 		{
 			| Ok(()) => return self.write_str("Successfully changed log level").await,
-			| Err(e) =>
-				return Err!("Failed to modify and reload the global tracing log level: {e}"),
+			| Err(e) => {
+				return Err!("Failed to modify and reload the global tracing log level: {e}");
+			},
 		}
 	}
 
@@ -529,6 +532,7 @@ pub(super) async fn force_set_room_state_from_server(
 	&self,
 	room_id: OwnedRoomId,
 	server_name: OwnedServerName,
+	at_event: Option<OwnedEventId>,
 ) -> Result {
 	if !self
 		.services
@@ -540,13 +544,18 @@ pub(super) async fn force_set_room_state_from_server(
 		return Err!("We are not participating in the room / we don't know about the room ID.");
 	}
 
-	let first_pdu = self
-		.services
-		.rooms
-		.timeline
-		.latest_pdu_in_room(&room_id)
-		.await
-		.map_err(|_| err!(Database("Failed to find the latest PDU in database")))?;
+	let at_event_id = match at_event {
+		| Some(event_id) => event_id,
+		| None => self
+			.services
+			.rooms
+			.timeline
+			.latest_pdu_in_room(&room_id)
+			.await
+			.map_err(|_| err!(Database("Failed to find the latest PDU in database")))?
+			.event_id
+			.clone(),
+	};
 
 	let room_version = self.services.rooms.state.get_room_version(&room_id).await?;
 
@@ -557,7 +566,7 @@ pub(super) async fn force_set_room_state_from_server(
 		.sending
 		.send_federation_request(&server_name, get_room_state::v1::Request {
 			room_id: room_id.clone(),
-			event_id: first_pdu.event_id.clone(),
+			event_id: at_event_id,
 		})
 		.await?;
 

src/admin/debug/mod.rs

@@ -177,6 +177,9 @@ pub(super) enum DebugCommand {
 		room_id: OwnedRoomId,
 		/// The server we will use to query the room state for
 		server_name: OwnedServerName,
+		/// The event ID of the latest known PDU in the room. Will be found
+		/// automatically if not provided.
+		event_id: Option<OwnedEventId>,
 	},
 
 	/// - Runs a server name through conduwuit's true destination resolution

src/admin/query/room_timeline.rs

@@ -31,7 +31,7 @@ pub(super) async fn last(&self, room_id: OwnedRoomOrAliasId) -> Result {
 		.services
 		.rooms
 		.timeline
-		.last_timeline_count(None, &room_id)
+		.last_timeline_count(&room_id)
 		.await?;
 
 	self.write_str(&format!("{result:#?}")).await
@@ -52,7 +52,7 @@ pub(super) async fn pdus(
 		.services
 		.rooms
 		.timeline
-		.pdus_rev(None, &room_id, from)
+		.pdus_rev(&room_id, from)
 		.try_take(limit.unwrap_or(3))
 		.try_collect()
 		.await?;

src/admin/room/moderation.rs

@@ -6,6 +6,7 @@ use conduwuit::{
 	warn,
 };
 use futures::StreamExt;
+use futures::FutureExt;
 use ruma::{OwnedRoomId, OwnedRoomOrAliasId, RoomAliasId, RoomId, RoomOrAliasId};
 
 use crate::{admin_command, admin_command_dispatch, get_room_info};
@@ -155,7 +156,10 @@ async fn ban_room(&self, room: OwnedRoomOrAliasId) -> Result {
 			 evicting admins too)",
 		);
 
-		if let Err(e) = leave_room(self.services, user_id, &room_id, None).await {
+		if let Err(e) = leave_room(self.services, user_id, &room_id, None)
+			.boxed()
+			.await
+		{
 			warn!("Failed to leave room: {e}");
 		}
 
@@ -323,7 +327,10 @@ async fn ban_list_of_rooms(&self) -> Result {
 				 evicting admins too)",
 			);
 
-			if let Err(e) = leave_room(self.services, user_id, &room_id, None).await {
+			if let Err(e) = leave_room(self.services, user_id, &room_id, None)
+				.boxed()
+				.await
+			{
 				warn!("Failed to leave room: {e}");
 			}
 

src/admin/user/commands.rs

@@ -9,6 +9,7 @@ use conduwuit::{
 };
 use conduwuit_api::client::{leave_all_rooms, update_avatar_url, update_displayname};
 use futures::StreamExt;
+use futures::FutureExt;
 use ruma::{
 	OwnedEventId, OwnedRoomId, OwnedRoomOrAliasId, OwnedUserId, UserId,
 	events::{
@@ -655,7 +656,9 @@ pub(super) async fn force_leave_room(
 		return Err!("{user_id} is not joined in the room");
 	}
 
-	leave_room(self.services, &user_id, &room_id, None).await?;
+	leave_room(self.services, &user_id, &room_id, None)
+		.boxed()
+		.await?;
 
 	self.write_str(&format!("{user_id} has left {room_id}.",))
 		.await

src/api/client/account.rs

@@ -763,7 +763,9 @@ pub(crate) async fn deactivate_route(
 	super::update_displayname(&services, sender_user, None, &all_joined_rooms).await;
 	super::update_avatar_url(&services, sender_user, None, None, &all_joined_rooms).await;
 
-	full_user_deactivate(&services, sender_user, &all_joined_rooms).await?;
+	full_user_deactivate(&services, sender_user, &all_joined_rooms)
+		.boxed()
+		.await?;
 
 	info!("User {sender_user} deactivated their account.");
 
@@ -915,7 +917,9 @@ pub async fn full_user_deactivate(
 		}
 	}
 
-	super::leave_all_rooms(services, user_id).await;
+	super::leave_all_rooms(services, user_id)
+		.boxed()
+		.await;
 
 	Ok(())
 }

src/api/client/context.rs

@@ -84,11 +84,25 @@ pub(crate) async fn get_context_route(
 
 	let base_event = ignored_filter(&services, (base_count, base_pdu), sender_user);
 
+	// PDUs are used to get seen user IDs and then returned in response.
+
 	let events_before = services
 		.rooms
 		.timeline
-		.pdus_rev(Some(sender_user), room_id, Some(base_count))
+		.pdus_rev(room_id, Some(base_count))
 		.ignore_err()
+		.then(async |mut pdu| {
+			pdu.1.set_unsigned(Some(sender_user));
+			if let Err(e) = services
+				.rooms
+				.pdu_metadata
+				.add_bundled_aggregations_to_pdu(sender_user, &mut pdu.1)
+				.await
+			{
+				debug_warn!("Failed to add bundled aggregations: {e}");
+			}
+			pdu
+		})
 		.ready_filter_map(|item| event_filter(item, filter))
 		.wide_filter_map(|item| ignored_filter(&services, item, sender_user))
 		.wide_filter_map(|item| visibility_filter(&services, item, sender_user))
@@ -98,8 +112,20 @@ pub(crate) async fn get_context_route(
 	let events_after = services
 		.rooms
 		.timeline
-		.pdus(Some(sender_user), room_id, Some(base_count))
+		.pdus(room_id, Some(base_count))
 		.ignore_err()
+		.then(async |mut pdu| {
+			pdu.1.set_unsigned(Some(sender_user));
+			if let Err(e) = services
+				.rooms
+				.pdu_metadata
+				.add_bundled_aggregations_to_pdu(sender_user, &mut pdu.1)
+				.await
+			{
+				debug_warn!("Failed to add bundled aggregations: {e}");
+			}
+			pdu
+		})
 		.ready_filter_map(|item| event_filter(item, filter))
 		.wide_filter_map(|item| ignored_filter(&services, item, sender_user))
 		.wide_filter_map(|item| visibility_filter(&services, item, sender_user))

src/api/client/membership.rs

@@ -114,7 +114,9 @@ async fn banned_room_check(
 					.collect()
 					.await;
 
-				full_user_deactivate(services, user_id, &all_joined_rooms).await?;
+				full_user_deactivate(services, user_id, &all_joined_rooms)
+					.boxed()
+					.await?;
 			}
 
 			return Err!(Request(Forbidden("This room is banned on this homeserver.")));
@@ -153,7 +155,9 @@ async fn banned_room_check(
 					.collect()
 					.await;
 
-				full_user_deactivate(services, user_id, &all_joined_rooms).await?;
+				full_user_deactivate(services, user_id, &all_joined_rooms)
+					.boxed()
+					.await?;
 			}
 
 			return Err!(Request(Forbidden("This remote server is banned on this homeserver.")));
@@ -259,6 +263,7 @@ pub(crate) async fn join_room_by_id_or_alias_route(
 				room_id.server_name(),
 				client,
 			)
+			.boxed()
 			.await?;
 
 			let mut servers = body.via.clone();
@@ -478,6 +483,7 @@ pub(crate) async fn leave_room_route(
 	body: Ruma<leave_room::v3::Request>,
 ) -> Result<leave_room::v3::Response> {
 	leave_room(&services, body.sender_user(), &body.room_id, body.reason.clone())
+		.boxed()
 		.await
 		.map(|()| leave_room::v3::Response::new())
 }
@@ -1243,6 +1249,7 @@ async fn join_room_by_id_helper_remote(
 		services.rooms.timeline.get_pdu(event_id).await.ok()
 	};
 
+	debug!("running stateres check on send_join parsed PDU");
 	let auth_check = state_res::event_auth::auth_check(
 		&state_res::RoomVersion::new(&room_version_id)?,
 		&parsed_join_pdu,
@@ -1792,7 +1799,10 @@ pub async fn leave_all_rooms(services: &Services, user_id: &UserId) {
 
 	for room_id in all_rooms {
 		// ignore errors
-		if let Err(e) = leave_room(services, user_id, &room_id, None).await {
+		if let Err(e) = leave_room(services, user_id, &room_id, None)
+			.boxed()
+			.await
+		{
 			warn!(%user_id, "Failed to leave {room_id} remotely: {e}");
 		}
 

src/api/client/message.rs

@@ -2,7 +2,7 @@ use core::panic;
 
 use axum::extract::State;
 use conduwuit::{
-	Err, Result, at,
+	Err, Result, at, debug_warn,
 	matrix::{
 		Event,
 		pdu::{PduCount, PduEvent},
@@ -114,14 +114,14 @@ pub(crate) async fn get_message_events_route(
 		| Direction::Forward => services
 			.rooms
 			.timeline
-			.pdus(Some(sender_user), room_id, Some(from))
+			.pdus(room_id, Some(from))
 			.ignore_err()
 			.boxed(),
 
 		| Direction::Backward => services
 			.rooms
 			.timeline
-			.pdus_rev(Some(sender_user), room_id, Some(from))
+			.pdus_rev(room_id, Some(from))
 			.ignore_err()
 			.boxed(),
 	};
@@ -132,6 +132,18 @@ pub(crate) async fn get_message_events_route(
 		.wide_filter_map(|item| ignored_filter(&services, item, sender_user))
 		.wide_filter_map(|item| visibility_filter(&services, item, sender_user))
 		.take(limit)
+		.then(async |mut pdu| {
+			pdu.1.set_unsigned(Some(sender_user));
+			if let Err(e) = services
+				.rooms
+				.pdu_metadata
+				.add_bundled_aggregations_to_pdu(sender_user, &mut pdu.1)
+				.await
+			{
+				debug_warn!("Failed to add bundled aggregations: {e}");
+			}
+			pdu
+		})
 		.collect()
 		.await;
 

src/api/client/relations.rs

@@ -1,6 +1,6 @@
 use axum::extract::State;
 use conduwuit::{
-	Result, at,
+	Result, at, debug_warn,
 	matrix::pdu::PduCount,
 	utils::{IterStream, ReadyExt, result::FlatOk, stream::WidebandExt},
 };
@@ -149,6 +149,17 @@ async fn paginate_relations_with_filter(
 		.ready_take_while(|(count, _)| Some(*count) != to)
 		.wide_filter_map(|item| visibility_filter(services, sender_user, item))
 		.take(limit)
+		.then(async |mut pdu| {
+			if let Err(e) = services
+				.rooms
+				.pdu_metadata
+				.add_bundled_aggregations_to_pdu(sender_user, &mut pdu.1)
+				.await
+			{
+				debug_warn!("Failed to add bundled aggregations to relation: {e}");
+			}
+			pdu
+		})
 		.collect()
 		.await;
 
@@ -172,6 +183,10 @@ async fn paginate_relations_with_filter(
 	})
 }
 
+// TODO: Can we move the visibility filter lower down, to avoid checking events
+// that won't be sent? At the moment this also results in getting events that
+// appear to have no relation because intermediaries are not visible to the
+// user.
 async fn visibility_filter(
 	services: &Services,
 	sender_user: &UserId,

src/api/client/room/event.rs

@@ -1,5 +1,5 @@
 use axum::extract::State;
-use conduwuit::{Err, Event, Result, err};
+use conduwuit::{Err, Event, Result, debug_warn, err};
 use futures::{FutureExt, TryFutureExt, future::try_join};
 use ruma::api::client::room::get_room_event;
 
@@ -38,7 +38,16 @@ pub(crate) async fn get_room_event_route(
 		"Fetched PDU must match requested"
 	);
 
-	event.add_age().ok();
+	if let Err(e) = services
+		.rooms
+		.pdu_metadata
+		.add_bundled_aggregations_to_pdu(body.sender_user(), &mut event)
+		.await
+	{
+		debug_warn!("Failed to add bundled aggregations to event: {e}");
+	}
+
+	event.set_unsigned(body.sender_user.as_deref());
 
 	Ok(get_room_event::v3::Response { event: event.into_room_event() })
 }

src/api/client/room/initial_sync.rs

@@ -1,6 +1,6 @@
 use axum::extract::State;
 use conduwuit::{
-	Err, PduEvent, Result, at,
+	Err, PduEvent, Result, at, debug_warn,
 	utils::{BoolExt, stream::TryTools},
 };
 use futures::TryStreamExt;
@@ -25,12 +25,28 @@ pub(crate) async fn room_initial_sync_route(
 		return Err!(Request(Forbidden("No room preview available.")));
 	}
 
+	// Events are returned in body
+
 	let limit = LIMIT_MAX;
 	let events: Vec<_> = services
 		.rooms
 		.timeline
-		.pdus_rev(None, room_id, None)
+		.pdus_rev(room_id, None)
 		.try_take(limit)
+		.and_then(async |mut pdu| {
+			pdu.1.set_unsigned(body.sender_user.as_deref());
+			if let Some(sender_user) = body.sender_user.as_deref() {
+				if let Err(e) = services
+					.rooms
+					.pdu_metadata
+					.add_bundled_aggregations_to_pdu(sender_user, &mut pdu.1)
+					.await
+				{
+					debug_warn!("Failed to add bundled aggregations: {e}");
+				}
+			}
+			Ok(pdu)
+		})
 		.try_collect()
 		.await?;
 

src/api/client/search.rs

@@ -2,7 +2,7 @@ use std::collections::BTreeMap;
 
 use axum::extract::State;
 use conduwuit::{
-	Err, Result, at, is_true,
+	Err, Result, at, debug_warn, is_true,
 	matrix::pdu::PduEvent,
 	result::FlatOk,
 	utils::{IterStream, stream::ReadyExt},
@@ -144,6 +144,17 @@ async fn category_room_events(
 		.map(at!(2))
 		.flatten()
 		.stream()
+		.then(|mut pdu| async {
+			if let Err(e) = services
+				.rooms
+				.pdu_metadata
+				.add_bundled_aggregations_to_pdu(sender_user, &mut pdu)
+				.await
+			{
+				debug_warn!("Failed to add bundled aggregations to search result: {e}");
+			}
+			pdu
+		})
 		.map(PduEvent::into_room_event)
 		.map(|result| SearchResult {
 			rank: None,

src/api/client/space.rs

@@ -121,7 +121,9 @@ where
 					.map(|(key, val)| (key, val.collect()))
 					.collect();
 
-				if !populate {
+				if populate {
+					rooms.push(summary_to_chunk(summary.clone()));
+				} else {
 					children = children
 						.iter()
 						.rev()
@@ -144,10 +146,8 @@ where
 						.collect();
 				}
 
-				if populate {
-					rooms.push(summary_to_chunk(summary.clone()));
-				} else if queue.is_empty() && children.is_empty() {
-					return Err!(Request(InvalidParam("Room IDs in token were not found.")));
+				if !populate && queue.is_empty() && children.is_empty() {
+					break;
 				}
 
 				parents.insert(current_room.clone());

src/api/client/state.rs

@@ -6,6 +6,7 @@ use conduwuit::{
 };
 use conduwuit_service::Services;
 use futures::TryStreamExt;
+use futures::FutureExt;
 use ruma::{
 	OwnedEventId, RoomId, UserId,
 	api::client::state::{get_state_events, get_state_events_for_key, send_state_event},
@@ -59,6 +60,7 @@ pub(crate) async fn send_state_event_for_empty_key_route(
 	body: Ruma<send_state_event::v3::Request>,
 ) -> Result<RumaResponse<send_state_event::v3::Response>> {
 	send_state_event_for_key_route(State(services), body)
+		.boxed()
 		.await
 		.map(RumaResponse)
 }

src/api/client/sync/mod.rs

@@ -3,7 +3,7 @@ mod v4;
 mod v5;
 
 use conduwuit::{
-	Error, PduCount, Result,
+	Error, PduCount, Result, debug_warn,
 	matrix::pdu::PduEvent,
 	utils::stream::{BroadbandExt, ReadyExt, TryIgnore},
 };
@@ -31,11 +31,7 @@ async fn load_timeline(
 	next_batch: Option<PduCount>,
 	limit: usize,
 ) -> Result<(Vec<(PduCount, PduEvent)>, bool), Error> {
-	let last_timeline_count = services
-		.rooms
-		.timeline
-		.last_timeline_count(Some(sender_user), room_id)
-		.await?;
+	let last_timeline_count = services.rooms.timeline.last_timeline_count(room_id).await?;
 
 	if last_timeline_count <= roomsincecount {
 		return Ok((Vec::new(), false));
@@ -44,10 +40,25 @@ async fn load_timeline(
 	let non_timeline_pdus = services
 		.rooms
 		.timeline
-		.pdus_rev(Some(sender_user), room_id, None)
+		.pdus_rev(room_id, None)
 		.ignore_err()
 		.ready_skip_while(|&(pducount, _)| pducount > next_batch.unwrap_or_else(PduCount::max))
-		.ready_take_while(|&(pducount, _)| pducount > roomsincecount);
+		.ready_take_while(|&(pducount, _)| pducount > roomsincecount)
+		.map(move |mut pdu| {
+			pdu.1.set_unsigned(Some(sender_user));
+			pdu
+		})
+		.then(async move |mut pdu| {
+			if let Err(e) = services
+				.rooms
+				.pdu_metadata
+				.add_bundled_aggregations_to_pdu(sender_user, &mut pdu.1)
+				.await
+			{
+				debug_warn!("Failed to add bundled aggregations: {e}");
+			}
+			pdu
+		});
 
 	// Take the last events for the timeline
 	pin_mut!(non_timeline_pdus);

src/api/client/sync/v3.rs

@@ -1009,8 +1009,6 @@ async fn calculate_state_incremental<'a>(
 ) -> Result<StateChanges> {
 	let since_shortstatehash = since_shortstatehash.unwrap_or(current_shortstatehash);
 
-	let state_changed = since_shortstatehash != current_shortstatehash;
-
 	let encrypted_room = services
 		.rooms
 		.state_accessor
@@ -1042,7 +1040,7 @@ async fn calculate_state_incremental<'a>(
 		})
 		.into();
 
-	let state_diff_ids: OptionFuture<_> = (!full_state && state_changed)
+	let state_diff_ids: OptionFuture<_> = (!full_state)
 		.then(|| {
 			StreamExt::into_future(
 				services
@@ -1191,7 +1189,7 @@ async fn calculate_heroes(
 	services
 		.rooms
 		.timeline
-		.all_pdus(sender_user, room_id)
+		.all_pdus(room_id)
 		.ready_filter(|(_, pdu)| pdu.kind == RoomMember)
 		.fold_default(|heroes: Vec<_>, (_, pdu)| {
 			fold_hero(heroes, services, room_id, sender_user, pdu)

src/api/client/threads.rs

@@ -1,6 +1,6 @@
 use axum::extract::State;
 use conduwuit::{
-	Result, at,
+	Result, at, debug_warn,
 	matrix::pdu::{PduCount, PduEvent},
 };
 use futures::StreamExt;
@@ -28,6 +28,8 @@ pub(crate) async fn get_threads_route(
 		.transpose()?
 		.unwrap_or_else(PduCount::max);
 
+	// TODO: user_can_see_event and set_unsigned should be at the same level /
+	// function, so unsigned is only set for seen events.
 	let threads: Vec<(PduCount, PduEvent)> = services
 		.rooms
 		.threads
@@ -42,6 +44,17 @@ pub(crate) async fn get_threads_route(
 				.await
 				.then_some((count, pdu))
 		})
+		.then(|(count, mut pdu)| async move {
+			if let Err(e) = services
+				.rooms
+				.pdu_metadata
+				.add_bundled_aggregations_to_pdu(body.sender_user(), &mut pdu)
+				.await
+			{
+				debug_warn!("Failed to add bundled aggregations to thread: {e}");
+			}
+			(count, pdu)
+		})
 		.collect()
 		.await;
 

src/api/server/backfill.rs

@@ -3,6 +3,7 @@ use std::cmp;
 use axum::extract::State;
 use conduwuit::{
 	PduCount, Result,
+	result::LogErr,
 	utils::{IterStream, ReadyExt, stream::TryTools},
 };
 use futures::{FutureExt, StreamExt, TryStreamExt};
@@ -62,7 +63,7 @@ pub(crate) async fn get_backfill_route(
 		pdus: services
 			.rooms
 			.timeline
-			.pdus_rev(None, &body.room_id, Some(from.saturating_add(1)))
+			.pdus_rev(&body.room_id, Some(from.saturating_add(1)))
 			.try_take(limit)
 			.try_filter_map(|(_, pdu)| async move {
 				Ok(services
@@ -72,6 +73,15 @@ pub(crate) async fn get_backfill_route(
 					.await
 					.then_some(pdu))
 			})
+			.and_then(async |mut pdu| {
+				// Strip the transaction ID, as that is private
+				pdu.remove_transaction_id().log_err().ok();
+				// Add age, as this is specified
+				pdu.add_age().log_err().ok();
+				// It's not clear if we should strip or add any more data, leave as is.
+				// In particular: Redaction?
+				Ok(pdu)
+			})
 			.try_filter_map(|pdu| async move {
 				Ok(services
 					.rooms

src/core/config/mod.rs

@@ -1155,7 +1155,7 @@ pub struct Config {
 	/// 3 to 5 = Statistics with possible performance impact.
 	/// 6 = All statistics.
 	///
-	/// default: 1
+	/// default: 3
 	#[serde(default = "default_rocksdb_stats_level")]
 	pub rocksdb_stats_level: u8,
 
@@ -1823,12 +1823,10 @@ pub struct Config {
 	pub stream_amplification: usize,
 
 	/// Number of sender task workers; determines sender parallelism. Default is
-	/// '0' which means the value is determined internally, likely matching the
-	/// number of tokio worker-threads or number of cores, etc. Override by
-	/// setting a non-zero value.
+	/// core count. Override by setting a different value.
 	///
-	/// default: 0
-	#[serde(default)]
+	/// default: core count
+	#[serde(default = "default_sender_workers")]
 	pub sender_workers: usize,
 
 	/// Enables listener sockets; can be set to false to disable listening. This
@@ -2059,45 +2057,47 @@ fn default_database_backups_to_keep() -> i16 { 1 }
 
 fn default_db_write_buffer_capacity_mb() -> f64 { 48.0 + parallelism_scaled_f64(4.0) }
 
-fn default_db_cache_capacity_mb() -> f64 { 128.0 + parallelism_scaled_f64(64.0) }
+fn default_db_cache_capacity_mb() -> f64 { 512.0 + parallelism_scaled_f64(512.0) }
 
-fn default_pdu_cache_capacity() -> u32 { parallelism_scaled_u32(10_000).saturating_add(100_000) }
+fn default_pdu_cache_capacity() -> u32 { parallelism_scaled_u32(50_000).saturating_add(500_000) }
 
 fn default_cache_capacity_modifier() -> f64 { 1.0 }
 
 fn default_auth_chain_cache_capacity() -> u32 {
-	parallelism_scaled_u32(10_000).saturating_add(100_000)
+	parallelism_scaled_u32(50_000).saturating_add(500_000)
 }
 
 fn default_shorteventid_cache_capacity() -> u32 {
-	parallelism_scaled_u32(50_000).saturating_add(100_000)
-}
-
-fn default_eventidshort_cache_capacity() -> u32 {
-	parallelism_scaled_u32(25_000).saturating_add(100_000)
-}
-
-fn default_eventid_pdu_cache_capacity() -> u32 {
-	parallelism_scaled_u32(25_000).saturating_add(100_000)
-}
-
-fn default_shortstatekey_cache_capacity() -> u32 {
-	parallelism_scaled_u32(10_000).saturating_add(100_000)
-}
-
-fn default_statekeyshort_cache_capacity() -> u32 {
-	parallelism_scaled_u32(10_000).saturating_add(100_000)
-}
-
-fn default_servernameevent_data_cache_capacity() -> u32 {
 	parallelism_scaled_u32(100_000).saturating_add(500_000)
 }
 
-fn default_stateinfo_cache_capacity() -> u32 { parallelism_scaled_u32(100) }
+fn default_eventidshort_cache_capacity() -> u32 {
+	parallelism_scaled_u32(100_000).saturating_add(500_000)
+}
 
-fn default_roomid_spacehierarchy_cache_capacity() -> u32 { parallelism_scaled_u32(1000) }
+fn default_eventid_pdu_cache_capacity() -> u32 {
+	parallelism_scaled_u32(50_000).saturating_add(500_000)
+}
 
-fn default_dns_cache_entries() -> u32 { 32768 }
+fn default_shortstatekey_cache_capacity() -> u32 {
+	parallelism_scaled_u32(50_000).saturating_add(500_000)
+}
+
+fn default_statekeyshort_cache_capacity() -> u32 {
+	parallelism_scaled_u32(50_000).saturating_add(500_000)
+}
+
+fn default_servernameevent_data_cache_capacity() -> u32 {
+	parallelism_scaled_u32(200_000).saturating_add(500_000)
+}
+
+fn default_stateinfo_cache_capacity() -> u32 {
+	parallelism_scaled_u32(500).clamp(100, 12000) }
+
+fn default_roomid_spacehierarchy_cache_capacity() -> u32 {
+	parallelism_scaled_u32(500).clamp(100, 12000) }
+
+fn default_dns_cache_entries() -> u32 { 327680 }
 
 fn default_dns_min_ttl() -> u64 { 60 * 180 }
 
@@ -2199,7 +2199,7 @@ fn default_typing_client_timeout_max_s() -> u64 { 45 }
 
 fn default_rocksdb_recovery_mode() -> u8 { 1 }
 
-fn default_rocksdb_log_level() -> String { "error".to_owned() }
+fn default_rocksdb_log_level() -> String { "info".to_owned() }
 
 fn default_rocksdb_log_time_to_roll() -> usize { 0 }
 
@@ -2231,7 +2231,7 @@ fn default_rocksdb_compression_level() -> i32 { 32767 }
 #[allow(clippy::doc_markdown)]
 fn default_rocksdb_bottommost_compression_level() -> i32 { 32767 }
 
-fn default_rocksdb_stats_level() -> u8 { 1 }
+fn default_rocksdb_stats_level() -> u8 { 3 }
 
 // I know, it's a great name
 #[must_use]
@@ -2286,14 +2286,13 @@ fn default_admin_log_capture() -> String {
 fn default_admin_room_tag() -> String { "m.server_notice".to_owned() }
 
 #[allow(clippy::as_conversions, clippy::cast_precision_loss)]
-fn parallelism_scaled_f64(val: f64) -> f64 { val * (sys::available_parallelism() as f64) }
+pub fn parallelism_scaled_f64(val: f64) -> f64 { val * (sys::available_parallelism() as f64) }
 
-fn parallelism_scaled_u32(val: u32) -> u32 {
-	let val = val.try_into().expect("failed to cast u32 to usize");
-	parallelism_scaled(val).try_into().unwrap_or(u32::MAX)
-}
+pub fn parallelism_scaled_u32(val: u32) -> u32 { val.saturating_mul(sys::available_parallelism() as u32) }
 
-fn parallelism_scaled(val: usize) -> usize { val.saturating_mul(sys::available_parallelism()) }
+pub fn parallelism_scaled_i32(val: i32) -> i32 { val.saturating_mul(sys::available_parallelism() as i32) }
+
+pub fn parallelism_scaled(val: usize) -> usize { val.saturating_mul(sys::available_parallelism()) }
 
 fn default_trusted_server_batch_size() -> usize { 256 }
 
@@ -2313,6 +2312,8 @@ fn default_stream_width_scale() -> f32 { 1.0 }
 
 fn default_stream_amplification() -> usize { 1024 }
 
+fn default_sender_workers() -> usize { parallelism_scaled(1) }
+
 fn default_client_receive_timeout() -> u64 { 75 }
 
 fn default_client_request_timeout() -> u64 { 180 }

src/core/matrix/pdu/unsigned.rs

@@ -1,11 +1,24 @@
-use std::collections::BTreeMap;
+use std::{borrow::Borrow, collections::BTreeMap};
 
 use ruma::MilliSecondsSinceUnixEpoch;
 use serde::Deserialize;
 use serde_json::value::{RawValue as RawJsonValue, Value as JsonValue, to_raw_value};
 
 use super::Pdu;
-use crate::{Result, err, implement, is_true};
+use crate::{Result, err, implement, is_true, result::LogErr};
+
+/// Set the `unsigned` field of the PDU using only information in the PDU.
+/// Some unsigned data is already set within the database (eg. prev events,
+/// threads). Once this is done, other data must be calculated from the database
+/// (eg. relations) This is for server-to-client events.
+/// Backfill handles this itself.
+#[implement(Pdu)]
+pub fn set_unsigned(&mut self, user_id: Option<&ruma::UserId>) {
+	if Some(self.sender.borrow()) != user_id {
+		self.remove_transaction_id().log_err().ok();
+	}
+	self.add_age().log_err().ok();
+}
 
 #[implement(Pdu)]
 pub fn remove_transaction_id(&mut self) -> Result {

src/core/matrix/state_res/event_auth.rs

@@ -13,6 +13,7 @@ use ruma::{
 		power_levels::RoomPowerLevelsEventContent,
 		third_party_invite::RoomThirdPartyInviteEventContent,
 	},
+	EventId,
 	int,
 	serde::{Base64, Raw},
 };
@@ -21,7 +22,6 @@ use serde::{
 	de::{Error as _, IgnoredAny},
 };
 use serde_json::{from_str as from_json_str, value::RawValue as RawJsonValue};
-
 use super::{
 	Error, Event, Result, StateEventType, StateKey, TimelineEventType,
 	power_levels::{
@@ -217,8 +217,9 @@ where
 	}
 
 	/*
-	// TODO: In the past this code caused problems federating with synapse, maybe this has been
-	// resolved already. Needs testing.
+	// TODO: In the past this code was commented as it caused problems with Synapse. This is no
+	// longer the case. This needs to be implemented.
+	// See also: https://github.com/ruma/ruma/pull/2064
 	//
 	// 2. Reject if auth_events
 	// a. auth_events cannot have duplicate keys since it's a BTree
@@ -241,20 +242,46 @@ where
 	}
 	*/
 
-	let (room_create_event, power_levels_event, sender_member_event) = join3(
-		fetch_state(&StateEventType::RoomCreate, ""),
-		fetch_state(&StateEventType::RoomPowerLevels, ""),
-		fetch_state(&StateEventType::RoomMember, sender.as_str()),
-	)
-	.await;
+	// let (room_create_event, power_levels_event, sender_member_event) = join3(
+	// 	fetch_state(&StateEventType::RoomCreate, ""),
+	// 	fetch_state(&StateEventType::RoomPowerLevels, ""),
+	// 	fetch_state(&StateEventType::RoomMember, sender.as_str()),
+	// )
+	// .await;
+
+	let room_create_event = fetch_state(&StateEventType::RoomCreate, "").await;
+	let power_levels_event = fetch_state(&StateEventType::RoomPowerLevels, "").await;
+	let sender_member_event = fetch_state(&StateEventType::RoomMember, sender.as_str()).await;
 
 	let room_create_event = match room_create_event {
 		| None => {
-			warn!("no m.room.create event in auth chain");
+			error!(
+				create_event = room_create_event.as_ref().map(Event::event_id).unwrap_or(<&EventId>::try_from("$unknown").unwrap()).as_str(),
+				power_levels = power_levels_event.as_ref().map(Event::event_id).unwrap_or(<&EventId>::try_from("$unknown").unwrap()).as_str(),
+				member_event = sender_member_event.as_ref().map(Event::event_id).unwrap_or(<&EventId>::try_from("$unknown").unwrap()).as_str(),
+				"no m.room.create event found for {} ({})!",
+				incoming_event.event_id().as_str(),
+				incoming_event.room_id().as_str()
+			);
 			return Ok(false);
 		},
 		| Some(e) => e,
 	};
+	// just re-check 1.2 to work around a bug
+	let Some(room_id_server_name) = incoming_event.room_id().server_name() else {
+		warn!("room ID has no servername");
+		return Ok(false);
+	};
+
+	if room_id_server_name != room_create_event.sender().server_name() {
+		warn!(
+			"servername of room ID origin ({}) does not match servername of m.room.create \
+			 sender ({})",
+			room_id_server_name,
+			room_create_event.sender().server_name()
+		);
+		return Ok(false);
+	}
 
 	// 3. If event does not have m.room.create in auth_events reject
 	if !incoming_event

src/core/matrix/state_res/mod.rs

@@ -609,7 +609,7 @@ where
 		let fetch_state = |ty: &StateEventType, key: &str| {
 			future::ready(auth_state.get(&ty.with_state_key(key)))
 		};
-
+		debug!("running auth check on {:?}", event.event_id());
 		let auth_result =
 			auth_check(room_version, &event, current_third_party.as_ref(), fetch_state).await;
 
@@ -726,8 +726,12 @@ where
 	Fut: Future<Output = Option<E>> + Send,
 	E: Event + Send + Sync,
 {
+	let mut room_id = None;
 	while let Some(sort_ev) = event {
-		debug!(event_id = sort_ev.event_id().as_str(), "mainline");
+		trace!(event_id = sort_ev.event_id().as_str(), "mainline");
+		if room_id.is_none() {
+			room_id = Some(sort_ev.room_id().to_owned());
+		}
 
 		let id = sort_ev.event_id();
 		if let Some(depth) = mainline_map.get(id) {
@@ -746,7 +750,7 @@ where
 			}
 		}
 	}
-	// Did not find a power level event so we default to zero
+	warn!("could not find a power event in the mainline map for {room_id:?}, defaulting to zero depth");
 	Ok(0)
 }
 

src/database/engine/cf_opts.rs

@@ -29,7 +29,7 @@ fn descriptor_cf_options(
 	set_table_options(&mut opts, &desc, cache)?;
 
 	opts.set_min_write_buffer_number(1);
-	opts.set_max_write_buffer_number(2);
+	opts.set_max_write_buffer_number(3);
 	opts.set_write_buffer_size(desc.write_size);
 
 	opts.set_target_file_size_base(desc.file_size);

src/database/engine/db_opts.rs

@@ -1,8 +1,6 @@
-use std::{cmp, convert::TryFrom};
-
-use conduwuit::{Config, Result, utils};
+use conduwuit::{Config, Result};
 use rocksdb::{Cache, DBRecoveryMode, Env, LogLevel, Options, statistics::StatsLevel};
-
+use conduwuit::config::{parallelism_scaled_i32, parallelism_scaled_u32};
 use super::{cf_opts::cache_size_f64, logger::handle as handle_log};
 
 /// Create database-wide options suitable for opening the database. This also
@@ -23,8 +21,8 @@ pub(crate) fn db_options(config: &Config, env: &Env, row_cache: &Cache) -> Resul
 	set_logging_defaults(&mut opts, config);
 
 	// Processing
-	opts.set_max_background_jobs(num_threads::<i32>(config)?);
-	opts.set_max_subcompactions(num_threads::<u32>(config)?);
+	opts.set_max_background_jobs(parallelism_scaled_i32(1));
+	opts.set_max_subcompactions(parallelism_scaled_u32(1));
 	opts.set_avoid_unnecessary_blocking_io(true);
 	opts.set_max_file_opening_threads(0);
 
@@ -126,15 +124,3 @@ fn set_logging_defaults(opts: &mut Options, config: &Config) {
 		opts.set_callback_logger(rocksdb_log_level, &handle_log);
 	}
 }
-
-fn num_threads<T: TryFrom<usize>>(config: &Config) -> Result<T> {
-	const MIN_PARALLELISM: usize = 2;
-
-	let requested = if config.rocksdb_parallelism_threads != 0 {
-		config.rocksdb_parallelism_threads
-	} else {
-		utils::available_parallelism()
-	};
-
-	utils::math::try_into::<T, usize>(cmp::max(MIN_PARALLELISM, requested))
-}

src/service/admin/mod.rs

@@ -4,7 +4,6 @@ mod execute;
 mod grant;
 
 use std::{
-	future::Future,
 	pin::Pin,
 	sync::{Arc, RwLock as StdRwLock, Weak},
 };
@@ -14,7 +13,7 @@ use conduwuit::{
 	Error, PduEvent, Result, Server, debug, err, error, error::default_log, pdu::PduBuilder,
 };
 pub use create::create_admin_room;
-use futures::{FutureExt, TryFutureExt};
+use futures::{Future, FutureExt, TryFutureExt};
 use loole::{Receiver, Sender};
 use ruma::{
 	OwnedEventId, OwnedRoomId, RoomId, UserId,

src/service/resolver/actual.rs

@@ -306,28 +306,25 @@ impl super::Service {
 
 	#[tracing::instrument(name = "srv", level = "debug", skip(self))]
 	async fn query_srv_record(&self, hostname: &'_ str) -> Result<Option<FedDest>> {
-		let hostnames =
-			[format!("_matrix-fed._tcp.{hostname}."), format!("_matrix._tcp.{hostname}.")];
+		self.services.server.check_running()?;
 
-		for hostname in hostnames {
-			self.services.server.check_running()?;
+		debug!("querying SRV for {hostname:?}");
 
-			debug!("querying SRV for {hostname:?}");
-			let hostname = hostname.trim_end_matches('.');
-			match self.resolver.resolver.srv_lookup(hostname).await {
-				| Err(e) => Self::handle_resolve_error(&e, hostname)?,
-				| Ok(result) => {
-					return Ok(result.iter().next().map(|result| {
-						FedDest::Named(
-							result.target().to_string().trim_end_matches('.').to_owned(),
-							format!(":{}", result.port())
-								.as_str()
-								.try_into()
-								.unwrap_or_else(|_| FedDest::default_port()),
-						)
-					}));
-				},
-			}
+		let hostname_suffix = format!("_matrix-fed._tcp.{hostname}.");
+		let hostname = hostname_suffix.trim_end_matches('.');
+		match self.resolver.resolver.srv_lookup(hostname).await {
+			| Err(e) => Self::handle_resolve_error(&e, hostname)?,
+			| Ok(result) => {
+				return Ok(result.iter().next().map(|result| {
+					FedDest::Named(
+						result.target().to_string().trim_end_matches('.').to_owned(),
+						format!(":{}", result.port())
+							.as_str()
+							.try_into()
+							.unwrap_or_else(|_| FedDest::default_port()),
+					)
+				}));
+			},
 		}
 
 		Ok(None)

src/service/rooms/event_handler/handle_outlier_pdu.rs

@@ -76,7 +76,7 @@ pub(super) async fn handle_outlier_pdu<'a>(
 		// 5. Reject "due to auth events" if can't get all the auth events or some of
 		//    the auth events are also rejected "due to auth events"
 		// NOTE: Step 5 is not applied anymore because it failed too often
-		debug!("Fetching auth events");
+		debug!("Fetching auth events for {}", incoming_pdu.event_id);
 		Box::pin(self.fetch_and_handle_outliers(
 			origin,
 			&incoming_pdu.auth_events,
@@ -88,12 +88,12 @@ pub(super) async fn handle_outlier_pdu<'a>(
 
 	// 6. Reject "due to auth events" if the event doesn't pass auth based on the
 	//    auth events
-	debug!("Checking based on auth events");
+	debug!("Checking {} based on auth events", incoming_pdu.event_id);
 	// Build map of auth events
 	let mut auth_events = HashMap::with_capacity(incoming_pdu.auth_events.len());
 	for id in &incoming_pdu.auth_events {
 		let Ok(auth_event) = self.services.timeline.get_pdu(id).await else {
-			warn!("Could not find auth event {id}");
+			warn!("Could not find auth event {id} for {}", incoming_pdu.event_id);
 			continue;
 		};
 
@@ -119,10 +119,7 @@ pub(super) async fn handle_outlier_pdu<'a>(
 	}
 
 	// The original create event must be in the auth events
-	if !matches!(
-		auth_events.get(&(StateEventType::RoomCreate, String::new().into())),
-		Some(_) | None
-	) {
+	if !auth_events.contains_key(&(StateEventType::RoomCreate, String::new().into())) {
 		return Err!(Request(InvalidParam("Incoming event refers to wrong create event.")));
 	}
 
@@ -131,6 +128,7 @@ pub(super) async fn handle_outlier_pdu<'a>(
 		ready(auth_events.get(&key))
 	};
 
+	debug!("running auth check to handle outlier pdu {:?}", incoming_pdu.event_id);
 	let auth_check = state_res::event_auth::auth_check(
 		&to_room_version(&room_version_id),
 		&incoming_pdu,

src/service/rooms/event_handler/upgrade_outlier_pdu.rs

@@ -1,12 +1,6 @@
 use std::{borrow::Borrow, collections::BTreeMap, iter::once, sync::Arc, time::Instant};
 
-use conduwuit::{
-	Err, Result, debug, debug_info, err, implement,
-	matrix::{EventTypeExt, PduEvent, StateKey, state_res},
-	trace,
-	utils::stream::{BroadbandExt, ReadyExt},
-	warn,
-};
+use conduwuit::{Err, Result, debug, debug_info, err, implement, matrix::{EventTypeExt, PduEvent, StateKey, state_res}, trace, utils::stream::{BroadbandExt, ReadyExt}, warn, info};
 use futures::{FutureExt, StreamExt, future::ready};
 use ruma::{CanonicalJsonValue, RoomId, ServerName, events::StateEventType};
 
@@ -44,7 +38,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 		return Err!(Request(InvalidParam("Event has been soft failed")));
 	}
 
-	debug!("Upgrading to timeline pdu");
+	debug!("Upgrading pdu {} from outlier to timeline pdu", incoming_pdu.event_id);
 	let timer = Instant::now();
 	let room_version_id = get_room_version_id(create_event)?;
 
@@ -52,7 +46,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 	//     backwards extremities doing all the checks in this list starting at 1.
 	//     These are not timeline events.
 
-	debug!("Resolving state at event");
+	debug!("Resolving state at event {}", incoming_pdu.event_id);
 	let mut state_at_incoming_event = if incoming_pdu.prev_events.len() == 1 {
 		self.state_at_incoming_degree_one(&incoming_pdu).await?
 	} else {
@@ -70,7 +64,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 		state_at_incoming_event.expect("we always set this to some above");
 	let room_version = to_room_version(&room_version_id);
 
-	debug!("Performing auth check");
+	debug!("Performing auth check to upgrade {}", incoming_pdu.event_id);
 	// 11. Check the auth of the event passes based on the state of the event
 	let state_fetch_state = &state_at_incoming_event;
 	let state_fetch = |k: StateEventType, s: StateKey| async move {
@@ -80,6 +74,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 		self.services.timeline.get_pdu(event_id).await.ok()
 	};
 
+	debug!("running auth check on {}", incoming_pdu.event_id);
 	let auth_check = state_res::event_auth::auth_check(
 		&room_version,
 		&incoming_pdu,
@@ -93,7 +88,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 		return Err!(Request(Forbidden("Event has failed auth check with state at the event.")));
 	}
 
-	debug!("Gathering auth events");
+	debug!("Gathering auth events for {}", incoming_pdu.event_id);
 	let auth_events = self
 		.services
 		.state
@@ -111,6 +106,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 		ready(auth_events.get(&key).cloned())
 	};
 
+	debug!("running auth check on {} with claimed state auth", incoming_pdu.event_id);
 	let auth_check = state_res::event_auth::auth_check(
 		&room_version,
 		&incoming_pdu,
@@ -121,7 +117,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 	.map_err(|e| err!(Request(Forbidden("Auth check failed: {e:?}"))))?;
 
 	// Soft fail check before doing state res
-	debug!("Performing soft-fail check");
+	debug!("Performing soft-fail check on {}", incoming_pdu.event_id);
 	let soft_fail = match (auth_check, incoming_pdu.redacts_id(&room_version_id)) {
 		| (false, _) => true,
 		| (true, None) => false,
@@ -145,7 +141,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 	let extremities: Vec<_> = self
 		.services
 		.state
-		.get_forward_extremities(room_id)
+		.get_forward_extremities(room_id, &state_lock)
 		.map(ToOwned::to_owned)
 		.ready_filter(|event_id| {
 			// Remove any that are referenced by this incoming event's prev_events
@@ -163,6 +159,8 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 		.collect()
 		.await;
 
+	if extremities.len() == 0 { info!("Retained zero extremities when upgrading outlier PDU to timeline PDU with {} previous events", incoming_pdu.prev_events.len()) }
+	
 	debug!(
 		"Retained {} extremities checked against {} prev_events",
 		extremities.len(),
@@ -218,7 +216,7 @@ pub(super) async fn upgrade_outlier_to_timeline_pdu(
 	// 14. Check if the event passes auth based on the "current state" of the room,
 	//     if not soft fail it
 	if soft_fail {
-		debug!("Soft failing event");
+		info!("Soft failing event {}", incoming_pdu.event_id);
 		let extremities = extremities.iter().map(Borrow::borrow);
 
 		self.services

src/service/rooms/pdu_metadata/bundled_aggregations.rs

@@ -0,0 +1,765 @@
+use conduwuit::{Event, PduEvent, Result, err};
+use ruma::{
+	EventId, RoomId, UserId,
+	api::Direction,
+	events::relation::{BundledMessageLikeRelations, BundledReference, ReferenceChunk},
+};
+
+use super::PdusIterItem;
+
+const MAX_BUNDLED_RELATIONS: usize = 50;
+
+impl super::Service {
+	/// Gets bundled aggregations for an event according to the Matrix
+	/// specification.
+	/// - m.replace relations are bundled to include the most recent replacement
+	///   event.
+	/// - m.reference relations are bundled to include a chunk of event IDs.
+	#[tracing::instrument(skip(self), level = "debug")]
+	pub async fn get_bundled_aggregations(
+		&self,
+		user_id: &UserId,
+		room_id: &RoomId,
+		event_id: &EventId,
+	) -> Result<Option<BundledMessageLikeRelations<Box<serde_json::value::RawValue>>>> {
+		let relations = self
+			.get_relations(
+				user_id,
+				room_id,
+				event_id,
+				conduwuit::PduCount::max(),
+				MAX_BUNDLED_RELATIONS,
+				0,
+				Direction::Backward,
+			)
+			.await;
+		// The relations database code still handles the basic unsigned data
+		// We don't want to recursively fetch relations
+
+		// TODO: Event visibility check
+		// TODO: ignored users?
+
+		if relations.is_empty() {
+			return Ok(None);
+		}
+
+		// Get the original event for validation of replacement events
+		let original_event = self.services.timeline.get_pdu(event_id).await?;
+
+		let mut replace_events = Vec::with_capacity(relations.len());
+		let mut reference_events = Vec::with_capacity(relations.len());
+
+		for relation in &relations {
+			let pdu = &relation.1;
+
+			let content = pdu.get_content_as_value();
+			if let Some(relates_to) = content.get("m.relates_to") {
+				// We don't check that the event relates back, because we assume the database is
+				// good.
+				if let Some(rel_type) = relates_to.get("rel_type") {
+					match rel_type.as_str() {
+						| Some("m.replace") => {
+							// Only consider valid replacements
+							if Self::is_valid_replacement_event(&original_event, pdu).await? {
+								replace_events.push(relation);
+							}
+						},
+						| Some("m.reference") => {
+							reference_events.push(relation);
+						},
+						| _ => {
+							// Ignore other relation types for now
+							// Threads are in the database but not handled here
+							// Other types are not specified AFAICT.
+						},
+					}
+				}
+			}
+		}
+
+		// If no relations to bundle, return None
+		if replace_events.is_empty() && reference_events.is_empty() {
+			return Ok(None);
+		}
+
+		let mut bundled = BundledMessageLikeRelations::new();
+
+		// Handle m.replace relations - find the most recent one
+		if !replace_events.is_empty() {
+			let most_recent_replacement = Self::find_most_recent_replacement(&replace_events)?;
+
+			// Convert the replacement event to the bundled format
+			if let Some(replacement_pdu) = most_recent_replacement {
+				// According to the Matrix spec, we should include the full event as raw JSON
+				let replacement_json = serde_json::to_string(replacement_pdu)
+					.map_err(|e| err!(Database("Failed to serialize replacement event: {e}")))?;
+				let raw_value = serde_json::value::RawValue::from_string(replacement_json)
+					.map_err(|e| err!(Database("Failed to create RawValue: {e}")))?;
+				bundled.replace = Some(Box::new(raw_value));
+			}
+		}
+
+		// Handle m.reference relations - collect event IDs
+		if !reference_events.is_empty() {
+			let reference_chunk = Self::build_reference_chunk(&reference_events)?;
+			if !reference_chunk.is_empty() {
+				bundled.reference = Some(Box::new(ReferenceChunk::new(reference_chunk)));
+			}
+		}
+
+		// TODO: Handle other relation types (m.annotation, etc.) when specified
+
+		Ok(Some(bundled))
+	}
+
+	/// Build reference chunk for m.reference bundled aggregations
+	fn build_reference_chunk(
+		reference_events: &[&PdusIterItem],
+	) -> Result<Vec<BundledReference>> {
+		let mut chunk = Vec::with_capacity(reference_events.len());
+
+		for relation in reference_events {
+			let pdu = &relation.1;
+
+			let reference_entry = BundledReference::new(pdu.event_id().to_owned());
+			chunk.push(reference_entry);
+		}
+
+		// Don't sort, order is unspecified
+
+		Ok(chunk)
+	}
+
+	/// Find the most recent replacement event based on origin_server_ts and
+	/// lexicographic event_id ordering
+	fn find_most_recent_replacement<'a>(
+		replacement_events: &'a [&'a PdusIterItem],
+	) -> Result<Option<&'a PduEvent>> {
+		if replacement_events.is_empty() {
+			return Ok(None);
+		}
+
+		let mut most_recent: Option<&PduEvent> = None;
+
+		// Jank, is there a better way to do this?
+		for relation in replacement_events {
+			let pdu = &relation.1;
+
+			match most_recent {
+				| None => {
+					most_recent = Some(pdu);
+				},
+				| Some(current_most_recent) => {
+					// Compare by origin_server_ts first
+					match pdu
+						.origin_server_ts()
+						.cmp(&current_most_recent.origin_server_ts())
+					{
+						| std::cmp::Ordering::Greater => {
+							most_recent = Some(pdu);
+						},
+						| std::cmp::Ordering::Equal => {
+							// If timestamps are equal, use lexicographic ordering of event_id
+							if pdu.event_id() > current_most_recent.event_id() {
+								most_recent = Some(pdu);
+							}
+						},
+						| std::cmp::Ordering::Less => {
+							// Keep current most recent
+						},
+					}
+				},
+			}
+		}
+
+		Ok(most_recent)
+	}
+
+	/// Adds bundled aggregations to a PDU's unsigned field
+	#[tracing::instrument(skip(self, pdu), level = "debug")]
+	pub async fn add_bundled_aggregations_to_pdu(
+		&self,
+		user_id: &UserId,
+		pdu: &mut PduEvent,
+	) -> Result<()> {
+		if pdu.is_redacted() {
+			return Ok(());
+		}
+
+		let bundled_aggregations = self
+			.get_bundled_aggregations(user_id, pdu.room_id(), pdu.event_id())
+			.await?;
+
+		if let Some(aggregations) = bundled_aggregations {
+			let aggregations_json = serde_json::to_value(aggregations)
+				.map_err(|e| err!(Database("Failed to serialize bundled aggregations: {e}")))?;
+
+			Self::add_bundled_aggregations_to_unsigned(pdu, aggregations_json)?;
+		}
+
+		Ok(())
+	}
+
+	/// Helper method to add bundled aggregations to a PDU's unsigned
+	/// field
+	fn add_bundled_aggregations_to_unsigned(
+		pdu: &mut PduEvent,
+		aggregations_json: serde_json::Value,
+	) -> Result<()> {
+		use serde_json::{
+			Map, Value as JsonValue,
+			value::{RawValue as RawJsonValue, to_raw_value},
+		};
+
+		let mut unsigned: Map<String, JsonValue> = pdu
+			.unsigned
+			.as_deref()
+			.map(RawJsonValue::get)
+			.map_or_else(|| Ok(Map::new()), serde_json::from_str)
+			.map_err(|e| err!(Database("Invalid unsigned in pdu event: {e}")))?;
+
+		let relations = unsigned
+			.entry("m.relations")
+			.or_insert_with(|| JsonValue::Object(Map::new()))
+			.as_object_mut()
+			.ok_or_else(|| err!(Database("m.relations is not an object")))?;
+
+		if let JsonValue::Object(aggregations_map) = aggregations_json {
+			for (rel_type, aggregation) in aggregations_map {
+				relations.insert(rel_type, aggregation);
+			}
+		}
+
+		pdu.unsigned = Some(to_raw_value(&unsigned)?);
+
+		Ok(())
+	}
+
+	/// Validates that an event is acceptable as a replacement for another event
+	/// See C/S spec "Validity of replacement events"
+	#[tracing::instrument(level = "debug")]
+	async fn is_valid_replacement_event(
+		original_event: &PduEvent,
+		replacement_event: &PduEvent,
+	) -> Result<bool> {
+		// 1. Same room_id
+		if original_event.room_id() != replacement_event.room_id() {
+			return Ok(false);
+		}
+
+		// 2. Same sender
+		if original_event.sender() != replacement_event.sender() {
+			return Ok(false);
+		}
+
+		// 3. Same type
+		if original_event.event_type() != replacement_event.event_type() {
+			return Ok(false);
+		}
+
+		// 4. Neither event should have a state_key property
+		if original_event.state_key().is_some() || replacement_event.state_key().is_some() {
+			return Ok(false);
+		}
+
+		// 5. Original event must not have rel_type of m.replace
+		let original_content = original_event.get_content_as_value();
+		if let Some(relates_to) = original_content.get("m.relates_to") {
+			if let Some(rel_type) = relates_to.get("rel_type") {
+				if rel_type.as_str() == Some("m.replace") {
+					return Ok(false);
+				}
+			}
+		}
+
+		// 6. Replacement event must have m.new_content property
+		// Skip this check for encrypted events, as m.new_content would be inside the
+		// encrypted payload
+		if replacement_event.event_type() != &ruma::events::TimelineEventType::RoomEncrypted {
+			let replacement_content = replacement_event.get_content_as_value();
+			if replacement_content.get("m.new_content").is_none() {
+				return Ok(false);
+			}
+		}
+
+		Ok(true)
+	}
+}
+
+#[cfg(test)]
+mod tests {
+	use conduwuit_core::pdu::{EventHash, PduEvent};
+	use ruma::{UInt, events::TimelineEventType, owned_event_id, owned_room_id, owned_user_id};
+	use serde_json::{Value as JsonValue, json, value::to_raw_value};
+
+	fn create_test_pdu(unsigned_content: Option<JsonValue>) -> PduEvent {
+		PduEvent {
+			event_id: owned_event_id!("$test:example.com"),
+			room_id: owned_room_id!("!test:example.com"),
+			sender: owned_user_id!("@test:example.com"),
+			origin_server_ts: UInt::try_from(1_234_567_890_u64).unwrap(),
+			kind: TimelineEventType::RoomMessage,
+			content: to_raw_value(&json!({"msgtype": "m.text", "body": "test"})).unwrap(),
+			state_key: None,
+			prev_events: vec![],
+			depth: UInt::from(1_u32),
+			auth_events: vec![],
+			redacts: None,
+			unsigned: unsigned_content.map(|content| to_raw_value(&content).unwrap()),
+			hashes: EventHash { sha256: "test_hash".to_owned() },
+			signatures: None,
+			origin: None,
+		}
+	}
+
+	fn create_bundled_aggregations() -> JsonValue {
+		json!({
+			"m.replace": {
+				"event_id": "$replace:example.com",
+				"origin_server_ts": 1_234_567_890,
+				"sender": "@replacer:example.com"
+			},
+			"m.reference": {
+				"count": 5,
+				"chunk": [
+					"$ref1:example.com",
+					"$ref2:example.com"
+				]
+			}
+		})
+	}
+
+	#[test]
+	fn test_add_bundled_aggregations_to_unsigned_no_existing_unsigned() {
+		let mut pdu = create_test_pdu(None);
+		let aggregations = create_bundled_aggregations();
+
+		let result = super::super::Service::add_bundled_aggregations_to_unsigned(
+			&mut pdu,
+			aggregations.clone(),
+		);
+		assert!(result.is_ok(), "Should succeed when no unsigned field exists");
+
+		assert!(pdu.unsigned.is_some(), "Unsigned field should be created");
+
+		let unsigned_str = pdu.unsigned.as_ref().unwrap().get();
+		let unsigned: JsonValue = serde_json::from_str(unsigned_str).unwrap();
+
+		assert!(unsigned.get("m.relations").is_some(), "m.relations should exist");
+		assert_eq!(
+			unsigned["m.relations"], aggregations,
+			"Relations should match the aggregations"
+		);
+	}
+
+	#[test]
+	fn test_add_bundled_aggregations_to_unsigned_overwrite_same_relation_type() {
+		let existing_unsigned = json!({
+			"m.relations": {
+				"m.replace": {
+					"event_id": "$old_replace:example.com",
+					"origin_server_ts": 1_111_111_111,
+					"sender": "@old_replacer:example.com"
+				}
+			}
+		});
+
+		let mut pdu = create_test_pdu(Some(existing_unsigned));
+		let new_aggregations = create_bundled_aggregations();
+
+		let result = super::super::Service::add_bundled_aggregations_to_unsigned(
+			&mut pdu,
+			new_aggregations.clone(),
+		);
+		assert!(result.is_ok(), "Should succeed when overwriting same relation type");
+
+		let unsigned_str = pdu.unsigned.as_ref().unwrap().get();
+		let unsigned: JsonValue = serde_json::from_str(unsigned_str).unwrap();
+
+		let relations = &unsigned["m.relations"];
+
+		assert_eq!(
+			relations["m.replace"], new_aggregations["m.replace"],
+			"m.replace should be updated"
+		);
+		assert_eq!(
+			relations["m.replace"]["event_id"], "$replace:example.com",
+			"Should have new event_id"
+		);
+
+		assert!(relations.get("m.reference").is_some(), "New m.reference should be added");
+	}
+
+	#[test]
+	fn test_add_bundled_aggregations_to_unsigned_preserve_other_unsigned_fields() {
+		// Test case: Other unsigned fields should be preserved
+		let existing_unsigned = json!({
+			"age": 98765,
+			"prev_content": {"msgtype": "m.text", "body": "old message"},
+			"redacted_because": {"event_id": "$redaction:example.com"},
+			"m.relations": {
+				"m.annotation": {"count": 1}
+			}
+		});
+
+		let mut pdu = create_test_pdu(Some(existing_unsigned));
+		let new_aggregations = json!({
+			"m.replace": {"event_id": "$new:example.com"}
+		});
+
+		let result = super::super::Service::add_bundled_aggregations_to_unsigned(
+			&mut pdu,
+			new_aggregations,
+		);
+		assert!(result.is_ok(), "Should succeed while preserving other fields");
+
+		let unsigned_str = pdu.unsigned.as_ref().unwrap().get();
+		let unsigned: JsonValue = serde_json::from_str(unsigned_str).unwrap();
+
+		// Verify all existing fields are preserved
+		assert_eq!(unsigned["age"], 98765, "age should be preserved");
+		assert!(unsigned.get("prev_content").is_some(), "prev_content should be preserved");
+		assert!(
+			unsigned.get("redacted_because").is_some(),
+			"redacted_because should be preserved"
+		);
+
+		// Verify relations were merged correctly
+		let relations = &unsigned["m.relations"];
+		assert!(
+			relations.get("m.annotation").is_some(),
+			"Existing m.annotation should be preserved"
+		);
+		assert!(relations.get("m.replace").is_some(), "New m.replace should be added");
+	}
+
+	#[test]
+	fn test_add_bundled_aggregations_to_unsigned_invalid_existing_unsigned() {
+		// Test case: Invalid JSON in existing unsigned should result in error
+		let mut pdu = create_test_pdu(None);
+		// Manually set invalid unsigned data
+		pdu.unsigned = Some(to_raw_value(&"invalid json").unwrap());
+
+		let aggregations = create_bundled_aggregations();
+		let result =
+			super::super::Service::add_bundled_aggregations_to_unsigned(&mut pdu, aggregations);
+
+		assert!(result.is_err(), "fails when existing unsigned is invalid");
+		// Should we ignore the error and overwrite anyway?
+	}
+
+	// Test helper function to create test PDU events
+	fn create_test_event(
+		event_id: &str,
+		room_id: &str,
+		sender: &str,
+		event_type: TimelineEventType,
+		content: &JsonValue,
+		state_key: Option<&str>,
+	) -> PduEvent {
+		PduEvent {
+			event_id: event_id.try_into().unwrap(),
+			room_id: room_id.try_into().unwrap(),
+			sender: sender.try_into().unwrap(),
+			origin_server_ts: UInt::try_from(1_234_567_890_u64).unwrap(),
+			kind: event_type,
+			content: to_raw_value(&content).unwrap(),
+			state_key: state_key.map(Into::into),
+			prev_events: vec![],
+			depth: UInt::from(1_u32),
+			auth_events: vec![],
+			redacts: None,
+			unsigned: None,
+			hashes: EventHash { sha256: "test_hash".to_owned() },
+			signatures: None,
+			origin: None,
+		}
+	}
+
+	/// Test that a valid replacement event passes validation
+	#[tokio::test]
+	async fn test_valid_replacement_event() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({"msgtype": "m.text", "body": "original message"}),
+			None,
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({
+				"msgtype": "m.text",
+				"body": "* edited message",
+				"m.new_content": {
+					"msgtype": "m.text",
+					"body": "edited message"
+				},
+				"m.relates_to": {
+					"rel_type": "m.replace",
+					"event_id": "$original:example.com"
+				}
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(result.unwrap(), "Valid replacement event should be accepted");
+	}
+
+	/// Test replacement event with different room ID is rejected
+	#[tokio::test]
+	async fn test_replacement_event_different_room() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room1:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({"msgtype": "m.text", "body": "original message"}),
+			None,
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room2:example.com", // Different room
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({
+				"msgtype": "m.text",
+				"body": "* edited message",
+				"m.new_content": {
+					"msgtype": "m.text",
+					"body": "edited message"
+				}
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(!result.unwrap(), "Different room ID should be rejected");
+	}
+
+	/// Test replacement event with different sender is rejected
+	#[tokio::test]
+	async fn test_replacement_event_different_sender() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room:example.com",
+			"@user1:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({"msgtype": "m.text", "body": "original message"}),
+			None,
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room:example.com",
+			"@user2:example.com", // Different sender
+			TimelineEventType::RoomMessage,
+			&json!({
+				"msgtype": "m.text",
+				"body": "* edited message",
+				"m.new_content": {
+					"msgtype": "m.text",
+					"body": "edited message"
+				}
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(!result.unwrap(), "Different sender should be rejected");
+	}
+
+	/// Test replacement event with different type is rejected
+	#[tokio::test]
+	async fn test_replacement_event_different_type() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({"msgtype": "m.text", "body": "original message"}),
+			None,
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomTopic, // Different event type
+			&json!({
+				"topic": "new topic",
+				"m.new_content": {
+					"topic": "new topic"
+				}
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(!result.unwrap(), "Different event type should be rejected");
+	}
+
+	/// Test replacement event with state key is rejected
+	#[tokio::test]
+	async fn test_replacement_event_with_state_key() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomName,
+			&json!({"name": "room name"}),
+			Some(""), // Has state key
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomName,
+			&json!({
+				"name": "new room name",
+				"m.new_content": {
+					"name": "new room name"
+				}
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(!result.unwrap(), "Event with state key should be rejected");
+	}
+
+	/// Test replacement of an event that is already a replacement is rejected
+	#[tokio::test]
+	async fn test_replacement_event_original_is_replacement() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({
+				"msgtype": "m.text",
+				"body": "* edited message",
+				"m.relates_to": {
+					"rel_type": "m.replace", // Original is already a replacement
+					"event_id": "$some_other:example.com"
+				}
+			}),
+			None,
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({
+				"msgtype": "m.text",
+				"body": "* edited again",
+				"m.new_content": {
+					"msgtype": "m.text",
+					"body": "edited again"
+				}
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(!result.unwrap(), "Replacement of replacement should be rejected");
+	}
+
+	/// Test replacement event missing m.new_content is rejected
+	#[tokio::test]
+	async fn test_replacement_event_missing_new_content() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({"msgtype": "m.text", "body": "original message"}),
+			None,
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomMessage,
+			&json!({
+				"msgtype": "m.text",
+				"body": "* edited message"
+				// Missing m.new_content
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(!result.unwrap(), "Missing m.new_content should be rejected");
+	}
+
+	/// Test encrypted replacement event without m.new_content is accepted
+	#[tokio::test]
+	async fn test_replacement_event_encrypted_missing_new_content_is_valid() {
+		let original = create_test_event(
+			"$original:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomEncrypted,
+			&json!({
+				"algorithm": "m.megolm.v1.aes-sha2",
+				"ciphertext": "encrypted_payload_base64",
+				"sender_key": "sender_key",
+				"session_id": "session_id"
+			}),
+			None,
+		);
+
+		let replacement = create_test_event(
+			"$replacement:example.com",
+			"!room:example.com",
+			"@user:example.com",
+			TimelineEventType::RoomEncrypted,
+			&json!({
+				"algorithm": "m.megolm.v1.aes-sha2",
+				"ciphertext": "encrypted_replacement_payload_base64",
+				"sender_key": "sender_key",
+				"session_id": "session_id",
+				"m.relates_to": {
+					"rel_type": "m.replace",
+					"event_id": "$original:example.com"
+				}
+				// No m.new_content in cleartext - this is valid for encrypted events
+			}),
+			None,
+		);
+
+		let result =
+			super::super::Service::is_valid_replacement_event(&original, &replacement).await;
+		assert!(result.is_ok(), "Validation should succeed");
+		assert!(
+			result.unwrap(),
+			"Encrypted replacement without cleartext m.new_content should be accepted"
+		);
+	}
+}

src/service/rooms/pdu_metadata/data.rs

@@ -3,7 +3,6 @@ use std::{mem::size_of, sync::Arc};
 use conduwuit::{
 	PduCount, PduEvent,
 	arrayvec::ArrayVec,
-	result::LogErr,
 	utils::{
 		ReadyExt,
 		stream::{TryIgnore, WidebandExt},
@@ -80,9 +79,7 @@ impl Data {
 
 			let mut pdu = self.services.timeline.get_pdu_from_id(&pdu_id).await.ok()?;
 
-			if pdu.sender != user_id {
-				pdu.remove_transaction_id().log_err().ok();
-			}
+			pdu.set_unsigned(Some(user_id));
 
 			Some((shorteventid, pdu))
 		})

src/service/rooms/pdu_metadata/mod.rs

@@ -1,3 +1,4 @@
+mod bundled_aggregations;
 mod data;
 use std::sync::Arc;
 

src/service/rooms/search/mod.rs

@@ -127,7 +127,12 @@ pub async fn search_pdus<'a>(
 				.then_some(pdu)
 		})
 		.skip(query.skip)
-		.take(query.limit);
+		.take(query.limit)
+		.map(move |mut pdu| {
+			pdu.set_unsigned(query.user_id);
+			// TODO: bundled aggregation
+			pdu
+		});
 
 	Ok((count, pdus))
 }

src/service/rooms/state/mod.rs

@@ -384,6 +384,7 @@ impl Service {
 	pub fn get_forward_extremities<'a>(
 		&'a self,
 		room_id: &'a RoomId,
+		_state_lock: &'a RoomMutexGuard,
 	) -> impl Stream<Item = &EventId> + Send + '_ {
 		let prefix = (room_id, Interfix);
 

src/service/rooms/threads/mod.rs

@@ -160,9 +160,7 @@ impl Service {
 				let mut pdu = self.services.timeline.get_pdu_from_id(&pdu_id).await.ok()?;
 				let pdu_id: PduId = pdu_id.into();
 
-				if pdu.sender != user_id {
-					pdu.remove_transaction_id().ok();
-				}
+				pdu.set_unsigned(Some(user_id));
 
 				Some((pdu_id.shorteventid, pdu))
 			});

src/service/rooms/timeline/data.rs

@@ -1,14 +1,11 @@
-use std::{borrow::Borrow, sync::Arc};
+use std::sync::Arc;
 
 use conduwuit::{
-	Err, PduCount, PduEvent, Result, at, err,
-	result::{LogErr, NotFound},
-	utils,
-	utils::stream::TryReadyExt,
+	Err, PduCount, PduEvent, Result, at, err, result::NotFound, utils, utils::stream::TryReadyExt,
 };
 use database::{Database, Deserialized, Json, KeyVal, Map};
 use futures::{FutureExt, Stream, TryFutureExt, TryStreamExt, future::select_ok, pin_mut};
-use ruma::{CanonicalJsonObject, EventId, OwnedUserId, RoomId, UserId, api::Direction};
+use ruma::{CanonicalJsonObject, EventId, OwnedUserId, RoomId, api::Direction};
 
 use super::{PduId, RawPduId};
 use crate::{Dep, rooms, rooms::short::ShortRoomId};
@@ -46,12 +43,8 @@ impl Data {
 	}
 
 	#[inline]
-	pub(super) async fn last_timeline_count(
-		&self,
-		sender_user: Option<&UserId>,
-		room_id: &RoomId,
-	) -> Result<PduCount> {
-		let pdus_rev = self.pdus_rev(sender_user, room_id, PduCount::max());
+	pub(super) async fn last_timeline_count(&self, room_id: &RoomId) -> Result<PduCount> {
+		let pdus_rev = self.pdus_rev(room_id, PduCount::max());
 
 		pin_mut!(pdus_rev);
 		let last_count = pdus_rev
@@ -65,12 +58,8 @@ impl Data {
 	}
 
 	#[inline]
-	pub(super) async fn latest_pdu_in_room(
-		&self,
-		sender_user: Option<&UserId>,
-		room_id: &RoomId,
-	) -> Result<PduEvent> {
-		let pdus_rev = self.pdus_rev(sender_user, room_id, PduCount::max());
+	pub(super) async fn latest_pdu_in_room(&self, room_id: &RoomId) -> Result<PduEvent> {
+		let pdus_rev = self.pdus_rev(room_id, PduCount::max());
 
 		pin_mut!(pdus_rev);
 		pdus_rev
@@ -223,7 +212,6 @@ impl Data {
 	/// order.
 	pub(super) fn pdus_rev<'a>(
 		&'a self,
-		user_id: Option<&'a UserId>,
 		room_id: &'a RoomId,
 		until: PduCount,
 	) -> impl Stream<Item = Result<PdusIterItem>> + Send + 'a {
@@ -233,14 +221,13 @@ impl Data {
 				self.pduid_pdu
 					.rev_raw_stream_from(&current)
 					.ready_try_take_while(move |(key, _)| Ok(key.starts_with(&prefix)))
-					.ready_and_then(move |item| Self::each_pdu(item, user_id))
+					.ready_and_then(Self::from_json_slice)
 			})
 			.try_flatten_stream()
 	}
 
 	pub(super) fn pdus<'a>(
 		&'a self,
-		user_id: Option<&'a UserId>,
 		room_id: &'a RoomId,
 		from: PduCount,
 	) -> impl Stream<Item = Result<PdusIterItem>> + Send + 'a {
@@ -250,21 +237,15 @@ impl Data {
 				self.pduid_pdu
 					.raw_stream_from(&current)
 					.ready_try_take_while(move |(key, _)| Ok(key.starts_with(&prefix)))
-					.ready_and_then(move |item| Self::each_pdu(item, user_id))
+					.ready_and_then(Self::from_json_slice)
 			})
 			.try_flatten_stream()
 	}
 
-	fn each_pdu((pdu_id, pdu): KeyVal<'_>, user_id: Option<&UserId>) -> Result<PdusIterItem> {
+	fn from_json_slice((pdu_id, pdu): KeyVal<'_>) -> Result<PdusIterItem> {
 		let pdu_id: RawPduId = pdu_id.into();
 
-		let mut pdu = serde_json::from_slice::<PduEvent>(pdu)?;
-
-		if Some(pdu.sender.borrow()) != user_id {
-			pdu.remove_transaction_id().log_err().ok();
-		}
-
-		pdu.add_age().log_err().ok();
+		let pdu = serde_json::from_slice::<PduEvent>(pdu)?;
 
 		Ok((pdu_id.pdu_count(), pdu))
 	}

src/service/rooms/timeline/mod.rs

@@ -165,7 +165,7 @@ impl Service {
 
 	#[tracing::instrument(skip(self), level = "debug")]
 	pub async fn first_item_in_room(&self, room_id: &RoomId) -> Result<(PduCount, PduEvent)> {
-		let pdus = self.pdus(None, room_id, None);
+		let pdus = self.pdus(room_id, None);
 
 		pin_mut!(pdus);
 		pdus.try_next()
@@ -175,16 +175,12 @@ impl Service {
 
 	#[tracing::instrument(skip(self), level = "debug")]
 	pub async fn latest_pdu_in_room(&self, room_id: &RoomId) -> Result<PduEvent> {
-		self.db.latest_pdu_in_room(None, room_id).await
+		self.db.latest_pdu_in_room(room_id).await
 	}
 
 	#[tracing::instrument(skip(self), level = "debug")]
-	pub async fn last_timeline_count(
-		&self,
-		sender_user: Option<&UserId>,
-		room_id: &RoomId,
-	) -> Result<PduCount> {
-		self.db.last_timeline_count(sender_user, room_id).await
+	pub async fn last_timeline_count(&self, room_id: &RoomId) -> Result<PduCount> {
+		self.db.last_timeline_count(room_id).await
 	}
 
 	/// Returns the `count` of this pdu's id.
@@ -545,6 +541,10 @@ impl Service {
 			| _ => {},
 		}
 
+		// CONCERN: If we receive events with a relation out-of-order, we never write
+		// their relation / thread. We need some kind of way to trigger when we receive
+		// this event, and potentially a way to rebuild the table entirely.
+
 		if let Ok(content) = pdu.get_content::<ExtractRelatesToEventId>() {
 			if let Ok(related_pducount) = self.get_pdu_count(&content.relates_to.event_id).await {
 				self.services
@@ -654,7 +654,7 @@ impl Service {
 		let prev_events: Vec<OwnedEventId> = self
 			.services
 			.state
-			.get_forward_extremities(room_id)
+			.get_forward_extremities(room_id, _mutex_lock)
 			.take(20)
 			.map(Into::into)
 			.collect()
@@ -698,6 +698,20 @@ impl Service {
 			.await
 			.saturating_add(uint!(1));
 
+		if state_key.is_none() {
+			if prev_events.is_empty() {
+				warn!("Timeline event had zero prev_events, something broke.");
+				return Err!(Request(Unknown("Timeline event had zero prev_events.")));
+			}
+			if depth.le(&uint!(2)) {
+				warn!(
+					"Had unsafe depth of {depth} in {room_id} when creating non-state event. \
+					 Bad!"
+				);
+				return Err!(Request(Unknown("Unsafe depth for non-state event.")));
+			}
+		};
+
 		let mut unsigned = unsigned.unwrap_or_default();
 
 		if let Some(state_key) = &state_key {
@@ -757,6 +771,7 @@ impl Service {
 			ready(auth_events.get(&key))
 		};
 
+		debug!("running auth check on new {} event by {} in {}", pdu.kind, pdu.sender, pdu.room_id);
 		let auth_check = state_res::auth_check(
 			&room_version,
 			&pdu,
@@ -996,34 +1011,30 @@ impl Service {
 	#[inline]
 	pub fn all_pdus<'a>(
 		&'a self,
-		user_id: &'a UserId,
 		room_id: &'a RoomId,
 	) -> impl Stream<Item = PdusIterItem> + Send + 'a {
-		self.pdus(Some(user_id), room_id, None).ignore_err()
+		self.pdus(room_id, None).ignore_err()
 	}
 
 	/// Reverse iteration starting at from.
 	#[tracing::instrument(skip(self), level = "debug")]
 	pub fn pdus_rev<'a>(
 		&'a self,
-		user_id: Option<&'a UserId>,
 		room_id: &'a RoomId,
 		until: Option<PduCount>,
 	) -> impl Stream<Item = Result<PdusIterItem>> + Send + 'a {
 		self.db
-			.pdus_rev(user_id, room_id, until.unwrap_or_else(PduCount::max))
+			.pdus_rev(room_id, until.unwrap_or_else(PduCount::max))
 	}
 
 	/// Forward iteration starting at from.
 	#[tracing::instrument(skip(self), level = "debug")]
 	pub fn pdus<'a>(
 		&'a self,
-		user_id: Option<&'a UserId>,
 		room_id: &'a RoomId,
 		from: Option<PduCount>,
 	) -> impl Stream<Item = Result<PdusIterItem>> + Send + 'a {
-		self.db
-			.pdus(user_id, room_id, from.unwrap_or_else(PduCount::min))
+		self.db.pdus(room_id, from.unwrap_or_else(PduCount::min))
 	}
 
 	/// Replace a PDU with the redacted form.
@@ -1142,7 +1153,7 @@ impl Service {
 			.boxed();
 
 		while let Some(ref backfill_server) = servers.next().await {
-			info!("Asking {backfill_server} for backfill");
+			info!("Asking {backfill_server} for backfill in {:?}", room_id.to_owned());
 			let response = self
 				.services
 				.sending
@@ -1170,7 +1181,7 @@ impl Service {
 			}
 		}
 
-		info!("No servers could backfill, but backfill was needed in room {room_id}");
+		warn!("No servers could backfill, but backfill was needed in room {room_id}");
 		Ok(())
 	}
 

src/service/sending/mod.rs

@@ -401,16 +401,10 @@ impl Service {
 
 fn num_senders(args: &crate::Args<'_>) -> usize {
 	const MIN_SENDERS: usize = 1;
-	// Limit the number of senders to the number of workers threads or number of
-	// cores, conservatively.
-	let max_senders = args
-		.server
-		.metrics
-		.num_workers()
-		.min(available_parallelism());
+	// Limit the maximum number of senders to the number of cores.
+	let max_senders = available_parallelism();
 
-	// If the user doesn't override the default 0, this is intended to then default
-	// to 1 for now as multiple senders is experimental.
+	// default is 4 senders. clamp between 1 and core count.
 	args.server
 		.config
 		.sender_workers

src/service/sending/sender.rs

@@ -781,7 +781,7 @@ impl Service {
 
 		for pdu in pdus {
 			// Redacted events are not notification targets (we don't send push for them)
-			if pdu.contains_unsigned_property("redacted_because", serde_json::Value::is_string) {
+			if pdu.is_redacted() {
 				continue;
 			}