Merge remote-tracking branch 'ginger/single-event-backfill' into illegal-car-mods

Partially borrowed from 85a84f93c7

conduwuit-example.toml

@@ -1074,7 +1074,7 @@
 # 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
@@ -1688,11 +1688,9 @@
 #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.

src/admin/debug/commands.rs

@@ -281,15 +281,8 @@ pub(super) async fn get_remote_pdu(
 				vec![(event_id, value, room_id)]
 			};
 
-			info!("Attempting to handle event ID {event_id} as backfilled PDU");
-			self.services
-				.rooms
-				.timeline
-				.backfill_pdu(&server, response.pdu)
-				.await?;
-
 			let text = serde_json::to_string_pretty(&json)?;
-			let msg = "Got PDU from specified server and handled as backfilled";
+			let msg = "Got PDU from specified server:";
 			write!(self, "{msg}. Event body:\n```json\n{text}\n```")
 		},
 	}

src/admin/query/room_timeline.rs

@@ -57,5 +57,5 @@ pub(super) async fn pdus(
 		.try_collect()
 		.await?;
 
-	self.write_str(&format!("{result:#?}")).await
+	self.write_str(&format!("```\n{result:#?}\n```")).await
 }

src/api/client/message.rs

@@ -35,7 +35,7 @@ use ruma::{
 };
 use tracing::warn;
 
-use super::utils::{count_to_token, parse_pagination_token as parse_token};
+use super::utils::{count_to_pagination_token, pagination_token_to_count as parse_token};
 use crate::Ruma;
 
 /// list of safe and common non-state events to ignore if the user is ignored
@@ -181,8 +181,8 @@ pub(crate) async fn get_message_events_route(
 		.collect();
 
 	Ok(get_message_events::v3::Response {
-		start: count_to_token(from),
-		end: next_token.map(count_to_token),
+		start: count_to_pagination_token(from),
+		end: next_token.map(count_to_pagination_token),
 		chunk,
 		state,
 	})

src/api/client/relations.rs

@@ -18,7 +18,7 @@ use ruma::{
 	events::{TimelineEventType, relation::RelationType},
 };
 
-use super::utils::{count_to_token, parse_pagination_token as parse_token};
+use super::utils::{count_to_pagination_token, pagination_token_to_count as parse_token};
 use crate::Ruma;
 
 /// # `GET /_matrix/client/r0/rooms/{roomId}/relations/{eventId}/{relType}/{eventType}`
@@ -193,7 +193,7 @@ async fn paginate_relations_with_filter(
 			| Direction::Forward => events.last(),
 			| Direction::Backward => events.first(),
 		}
-		.map(|(count, _)| count_to_token(*count))
+		.map(|(count, _)| count_to_pagination_token(*count))
 	} else {
 		None
 	};

src/api/client/room/event.rs

@@ -18,7 +18,7 @@ pub(crate) async fn get_room_event_route(
 	let event = services
 		.rooms
 		.timeline
-		.get_pdu(event_id)
+		.get_remote_pdu(room_id, event_id)
 		.map_err(|_| err!(Request(NotFound("Event {} not found.", event_id))));
 
 	let visible = services
@@ -33,11 +33,6 @@ pub(crate) async fn get_room_event_route(
 		return Err!(Request(Forbidden("You don't have permission to view this event.")));
 	}
 
-	debug_assert!(
-		event.event_id() == event_id && event.room_id() == room_id,
-		"Fetched PDU must match requested"
-	);
-
 	event.add_age().ok();
 
 	Ok(get_room_event::v3::Response { event: event.into_format() })

src/api/client/threads.rs

@@ -9,7 +9,7 @@ use conduwuit::{
 use futures::StreamExt;
 use ruma::{api::client::threads::get_threads, uint};
 
-use crate::Ruma;
+use crate::{Ruma, client::utils::pagination_token_to_count};
 
 /// # `GET /_matrix/client/r0/rooms/{roomId}/threads`
 pub(crate) async fn get_threads_route(
@@ -27,7 +27,7 @@ pub(crate) async fn get_threads_route(
 	let from: PduCount = body
 		.from
 		.as_deref()
-		.map(str::parse)
+		.map(pagination_token_to_count)
 		.transpose()?
 		.unwrap_or_else(PduCount::max);
 

src/api/client/utils.rs

@@ -1,28 +1,7 @@
-use conduwuit::{
-	Result, err,
-	matrix::pdu::{PduCount, ShortEventId},
-};
+use conduwuit::{Result, matrix::pdu::PduCount};
 
-/// Parse a pagination token, trying ShortEventId first, then falling back to
-/// PduCount
-pub(crate) fn parse_pagination_token(token: &str) -> Result<PduCount> {
-	// Try parsing as ShortEventId first
-	if let Ok(shorteventid) = token.parse::<ShortEventId>() {
-		// ShortEventId maps directly to a PduCount in our database
-		Ok(PduCount::Normal(shorteventid))
-	} else if let Ok(count) = token.parse::<u64>() {
-		// Fallback to PduCount for backwards compatibility
-		Ok(PduCount::Normal(count))
-	} else if let Ok(count) = token.parse::<i64>() {
-		// Also handle negative counts for backfilled events
-		Ok(PduCount::from_signed(count))
-	} else {
-		Err(err!(Request(InvalidParam("Invalid pagination token"))))
-	}
-}
+/// Parse a pagination token
+pub(crate) fn pagination_token_to_count(token: &str) -> Result<PduCount> { token.parse() }
 
-/// Convert a PduCount to a token string (using the underlying ShortEventId)
-pub(crate) fn count_to_token(count: PduCount) -> String {
-	// The PduCount's unsigned value IS the ShortEventId
-	count.into_unsigned().to_string()
-}
+/// Convert a PduCount to a token string
+pub(crate) fn count_to_pagination_token(count: PduCount) -> String { count.to_string() }

src/api/server/send.rs

@@ -3,9 +3,9 @@ use std::{collections::BTreeMap, net::IpAddr, time::Instant};
 use axum::extract::State;
 use axum_client_ip::InsecureClientIp;
 use conduwuit::{
-	Err, Error, Result, debug,
+	Err, Error, Result,
 	debug::INFO_SPAN_LEVEL,
-	debug_warn, err, error,
+	debug_warn, err, error, info,
 	result::LogErr,
 	trace,
 	utils::{
@@ -79,23 +79,14 @@ pub(crate) async fn send_transaction_message_route(
 	}
 
 	let txn_start_time = Instant::now();
+	let transaction_id = body.transaction_id.to_string();
 	trace!(
 		pdus = body.pdus.len(),
 		edus = body.edus.len(),
-		elapsed = ?txn_start_time.elapsed(),
-		id = ?body.transaction_id,
-		origin =?body.origin(),
-		"Starting txn",
+		id = transaction_id,
+		"Processing transaction",
 	);
 
-	let pdus = body
-		.pdus
-		.iter()
-		.stream()
-		.broad_then(|pdu| services.rooms.event_handler.parse_incoming_pdu(pdu))
-		.inspect_err(|e| debug_warn!("Could not parse PDU: {e}"))
-		.ready_filter_map(Result::ok);
-
 	let edus = body
 		.edus
 		.iter()
@@ -104,14 +95,29 @@ pub(crate) async fn send_transaction_message_route(
 		.filter_map(Result::ok)
 		.stream();
 
-	let results = handle(&services, &client, body.origin(), txn_start_time, pdus, edus).await?;
+	let pdus = body
+		.pdus
+		.iter()
+		.stream()
+		.broad_then(|pdu| services.rooms.event_handler.parse_incoming_pdu(pdu))
+		.inspect_err(|e| debug_warn!("Could not parse PDU: {e}"))
+		.ready_filter_map(Result::ok);
 
-	debug!(
+	trace!(
 		pdus = body.pdus.len(),
 		edus = body.edus.len(),
 		elapsed = ?txn_start_time.elapsed(),
-		id = ?body.transaction_id,
-		origin =?body.origin(),
+		id = transaction_id,
+		"Validated transaction",
+	);
+
+	let results = handle(&services, &client, body.origin(), txn_start_time, pdus, edus, &*transaction_id).await?;
+
+	info!(
+		pdus = body.pdus.len(),
+		edus = body.edus.len(),
+		elapsed = ?txn_start_time.elapsed(),
+		id = transaction_id,
 		"Finished txn",
 	);
 	for (id, result) in &results {
@@ -137,7 +143,13 @@ async fn handle(
 	started: Instant,
 	pdus: impl Stream<Item = Pdu> + Send,
 	edus: impl Stream<Item = Edu> + Send,
+	transaction_id: &str,
 ) -> Result<ResolvedMap> {
+	let handle_start = Instant::now();
+	edus.for_each_concurrent(automatic_width(), |edu| handle_edu(services, client, origin, edu))
+		.boxed()
+		.await;
+	let pdu_start = Instant::now();
 	// group pdus by room
 	let pdus = pdus
 		.collect()
@@ -148,13 +160,14 @@ async fn handle(
 				.collect()
 		})
 		.await;
-
+	let results_start = Instant::now();
 	// we can evaluate rooms concurrently
 	let results: ResolvedMap = pdus
 		.into_iter()
 		.try_stream()
 		.broad_and_then(|(room_id, pdus): (_, Vec<_>)| {
-			handle_room(services, client, origin, started, room_id, pdus.into_iter())
+			let count = pdus.len();
+			handle_room(services, client, origin, started, room_id, pdus.into_iter(), count, transaction_id)
 				.map_ok(Vec::into_iter)
 				.map_ok(IterStream::try_stream)
 		})
@@ -162,12 +175,14 @@ async fn handle(
 		.try_collect()
 		.boxed()
 		.await?;
-
-	// evaluate edus after pdus, at least for now.
-	edus.for_each_concurrent(automatic_width(), |edu| handle_edu(services, client, origin, edu))
-		.boxed()
-		.await;
-
+	let handle_stop = Instant::now();
+	info!(
+		edus = pdu_start.saturating_duration_since(handle_start).as_micros(),
+		pdus = results_start.saturating_duration_since(pdu_start).as_micros(),
+		handle_room = handle_stop.saturating_duration_since(results_start).as_micros(),
+		transaction_id = ?transaction_id,
+		"handled incoming transaction",
+	);
 	Ok(results)
 }
 
@@ -178,19 +193,35 @@ async fn handle_room(
 	txn_start_time: Instant,
 	room_id: OwnedRoomId,
 	pdus: impl Iterator<Item = Pdu> + Send,
+	count: usize,
+	transaction_id: &str,
 ) -> Result<Vec<(OwnedEventId, Result)>> {
+	let room_lock_start = Instant::now();
 	let _room_lock = services
 		.rooms
 		.event_handler
 		.mutex_federation
 		.lock(&room_id)
 		.await;
+	let room_lock_end = Instant::now();
 
 	let room_id = &room_id;
+	let mut n = 0;
 	pdus.try_stream()
 		.and_then(|(_, event_id, value)| async move {
 			services.server.check_running()?;
 			let pdu_start_time = Instant::now();
+			trace!(
+				%room_id,
+				%event_id,
+				transaction_id = ?transaction_id,
+				pdu = n + 1,
+				total = count,
+				room_lock_time = ?room_lock_end.saturating_duration_since(room_lock_start).as_micros(),
+				pdu_elapsed = ?pdu_start_time.elapsed(),
+				txn_elapsed = ?txn_start_time.elapsed(),
+				"Handling PDU",
+			);
 			let result = services
 				.rooms
 				.event_handler
@@ -198,11 +229,18 @@ async fn handle_room(
 				.await
 				.map(|_| ());
 
-			debug!(
+			info!(
+				%room_id,
+				%event_id,
+				transaction_id = ?transaction_id,
+				pdu = n + 1,
+				total = count,
+				room_lock_time = ?room_lock_end.saturating_duration_since(room_lock_start).as_micros(),
 				pdu_elapsed = ?pdu_start_time.elapsed(),
 				txn_elapsed = ?txn_start_time.elapsed(),
-				"Finished PDU {event_id}",
+				"Finished handling PDU",
 			);
+			n += 1;
 
 			Ok((event_id, result))
 		})

src/core/config/mod.rs

@@ -1247,7 +1247,7 @@ 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,
 
@@ -1929,12 +1929,10 @@ 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
@@ -2277,45 +2275,48 @@ 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(100_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)
-}
-
-fn default_shorteventid_cache_capacity() -> u32 {
 	parallelism_scaled_u32(50_000).saturating_add(100_000)
 }
 
+fn default_shorteventid_cache_capacity() -> u32 {
+	parallelism_scaled_u32(100_000).saturating_add(100_000)
+}
+
 fn default_eventidshort_cache_capacity() -> u32 {
-	parallelism_scaled_u32(25_000).saturating_add(100_000)
+	parallelism_scaled_u32(50_000).saturating_add(100_000)
 }
 
 fn default_eventid_pdu_cache_capacity() -> u32 {
-	parallelism_scaled_u32(25_000).saturating_add(100_000)
+	parallelism_scaled_u32(50_000).saturating_add(100_000)
 }
 
 fn default_shortstatekey_cache_capacity() -> u32 {
-	parallelism_scaled_u32(10_000).saturating_add(100_000)
+	parallelism_scaled_u32(50_000).saturating_add(100_000)
 }
 
 fn default_statekeyshort_cache_capacity() -> u32 {
-	parallelism_scaled_u32(10_000).saturating_add(100_000)
+	parallelism_scaled_u32(50_000).saturating_add(100_000)
 }
 
 fn default_servernameevent_data_cache_capacity() -> u32 {
-	parallelism_scaled_u32(100_000).saturating_add(500_000)
+	parallelism_scaled_u32(100_000).saturating_add(100_000)
 }
 
-fn default_stateinfo_cache_capacity() -> u32 { parallelism_scaled_u32(100) }
+fn default_stateinfo_cache_capacity() -> u32 {
+	parallelism_scaled_u32(500).clamp(100, 12000)
+}
 
-fn default_roomid_spacehierarchy_cache_capacity() -> u32 { parallelism_scaled_u32(1000) }
+fn default_roomid_spacehierarchy_cache_capacity() -> u32 {
+	parallelism_scaled_u32(500).clamp(100, 12000) }
 
-fn default_dns_cache_entries() -> u32 { 32768 }
+fn default_dns_cache_entries() -> u32 { 327680 }
 
 fn default_dns_min_ttl() -> u64 { 60 * 180 }
 
@@ -2421,7 +2422,7 @@ 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 }
 
@@ -2453,7 +2454,7 @@ 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]
@@ -2508,14 +2509,13 @@ 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 }
 
@@ -2535,6 +2535,8 @@ 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/count.rs

@@ -1,6 +1,10 @@
 #![allow(clippy::cast_possible_wrap, clippy::cast_sign_loss, clippy::as_conversions)]
 
-use std::{cmp::Ordering, fmt, fmt::Display, str::FromStr};
+use std::{
+	cmp::Ordering,
+	fmt::{self, Display},
+	str::FromStr,
+};
 
 use ruma::api::Direction;
 

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::{
@@ -149,8 +149,8 @@ where
 	for<'a> &'a E: Event + Send,
 {
 	debug!(
-		event_id = %incoming_event.event_id(),
-		event_type = ?incoming_event.event_type(),
+		event_id = format!("{}", incoming_event.event_id()),
+		event_type = format!("{}", incoming_event.event_type()),
 		"auth_check beginning"
 	);
 
@@ -219,7 +219,7 @@ where
 	/*
 	// 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
+	// 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
@@ -242,20 +242,44 @@ 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);
+	}
 
 	if incoming_event.room_id() != room_create_event.room_id() {
 		warn!("room_id of incoming event does not match room_id of m.room.create event");

src/core/matrix/state_res/mod.rs

@@ -733,8 +733,12 @@ 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");
+		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) {
@@ -753,7 +757,7 @@ 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/federation/execute.rs

@@ -1,10 +1,11 @@
-use std::{fmt::Debug, mem};
+use std::{
+	error::Error as _,
+	fmt::{Debug, Write},
+	mem,
+};
 
 use bytes::Bytes;
-use conduwuit::{
-	Err, Error, Result, debug, debug::INFO_SPAN_LEVEL, debug_error, debug_warn, err,
-	error::inspect_debug_log, implement, trace, utils::string::EMPTY,
-};
+use conduwuit::{Err, Error, Result, debug, debug::INFO_SPAN_LEVEL, err, error::inspect_debug_log, implement, trace, utils::string::EMPTY, info};
 use http::{HeaderValue, header::AUTHORIZATION};
 use ipaddress::IPAddress;
 use reqwest::{Client, Method, Request, Response, Url};
@@ -193,9 +194,9 @@ fn handle_error(
 ) -> Result {
 	if e.is_timeout() || e.is_connect() {
 		e = e.without_url();
-		debug_warn!("{e:?}");
+		trace!(?url, "network error while sending federation request: {e:?}");
 	} else if e.is_redirect() {
-		debug_error!(
+		trace!(
 			method = ?method,
 			url = ?url,
 			final_url = ?e.url(),
@@ -204,9 +205,17 @@ fn handle_error(
 			e,
 		);
 	} else {
-		debug_error!("{e:?}");
+		trace!(?url, "failed to send federation request: {e:?}");
 	}
 
+	let mut nice_error = "Request failed".to_owned();
+	let mut src = e.source();
+	while let Some(source) = src {
+		write!(nice_error, ": {source:?}").expect("writing to string should not fail");
+		src = source.source();
+	}
+	info!(nice_error, "Federation request error");
+
 	Err(e.into())
 }
 

src/service/presence/mod.rs

@@ -100,7 +100,7 @@ impl Service {
 	/// Pings the presence of the given user in the given room, setting the
 	/// specified state.
 	pub async fn ping_presence(&self, user_id: &UserId, new_state: &PresenceState) -> Result<()> {
-		const REFRESH_TIMEOUT: u64 = 60 * 1000;
+		const REFRESH_TIMEOUT: u64 = 60 * 1000 * 4;
 
 		let last_presence = self.db.get_presence(user_id).await;
 		let state_changed = match last_presence {

src/service/resolver/dns.rs

@@ -53,9 +53,9 @@ impl Resolver {
 		opts.cache_size = config.dns_cache_entries as usize;
 		opts.preserve_intermediates = true;
 		opts.negative_min_ttl = Some(Duration::from_secs(config.dns_min_ttl_nxdomain));
-		opts.negative_max_ttl = Some(Duration::from_secs(60 * 60 * 24 * 30));
+		opts.negative_max_ttl = Some(Duration::from_secs(60 * 60 * 24));
 		opts.positive_min_ttl = Some(Duration::from_secs(config.dns_min_ttl));
-		opts.positive_max_ttl = Some(Duration::from_secs(60 * 60 * 24 * 7));
+		opts.positive_max_ttl = Some(Duration::from_secs(60 * 60 * 24));
 		opts.timeout = Duration::from_secs(config.dns_timeout);
 		opts.attempts = config.dns_attempts as usize;
 		opts.try_tcp_on_error = config.dns_tcp_fallback;

src/service/rooms/event_handler/fetch_and_handle_outliers.rs

@@ -3,11 +3,7 @@ use std::{
 	time::Instant,
 };
 
-use conduwuit::{
-	Event, PduEvent, debug, debug_error, debug_warn, implement,
-	matrix::event::gen_event_id_canonical_json, trace, utils::continue_exponential_backoff_secs,
-	warn,
-};
+use conduwuit::{Event, PduEvent, debug, debug_error, implement, matrix::event::gen_event_id_canonical_json, trace, utils::continue_exponential_backoff_secs, warn, debug_warn};
 use ruma::{
 	CanonicalJsonValue, EventId, OwnedEventId, RoomId, ServerName,
 	api::federation::event::get_event,
@@ -79,7 +75,7 @@ where
 			{
 				// Exponential backoff
 				const MIN_DURATION: u64 = 60 * 2;
-				const MAX_DURATION: u64 = 60 * 60 * 8;
+				const MAX_DURATION: u64 = 60 * 60;
 				if continue_exponential_backoff_secs(
 					MIN_DURATION,
 					MAX_DURATION,

src/service/rooms/event_handler/handle_outlier_pdu.rs

@@ -122,11 +122,8 @@ where
 	}
 
 	// The original create event must be in the auth events
-	if !matches!(
-		auth_events.get(&(StateEventType::RoomCreate, String::new().into())),
-		Some(_) | None
-	) {
-		return Err!(Request(InvalidParam("Incoming event refers to wrong create event.")));
+	if !auth_events.contains_key(&(StateEventType::RoomCreate, String::new().into())) {
+		return Err!(Request(InvalidParam("Incoming event refers to wrong create event. event_id={event_id}")));
 	}
 
 	let state_fetch = |ty: &StateEventType, sk: &str| {

src/service/rooms/event_handler/handle_prev_pdu.rs

@@ -46,7 +46,7 @@ where
 	{
 		// Exponential backoff
 		const MIN_DURATION: u64 = 5 * 60;
-		const MAX_DURATION: u64 = 60 * 60 * 24;
+		const MAX_DURATION: u64 = 60 * 60;
 		if continue_exponential_backoff_secs(MIN_DURATION, MAX_DURATION, time.elapsed(), *tries) {
 			debug!(
 				?tries,

src/service/rooms/event_handler/upgrade_outlier_pdu.rs

@@ -5,7 +5,7 @@ use conduwuit::{
 	matrix::{Event, EventTypeExt, PduEvent, StateKey, state_res},
 	trace,
 	utils::stream::{BroadbandExt, ReadyExt},
-	warn,
+	warn
 };
 use futures::{FutureExt, StreamExt, future::ready};
 use ruma::{CanonicalJsonValue, RoomId, ServerName, events::StateEventType};
@@ -193,6 +193,8 @@ where
 		.collect()
 		.await;
 
+	if extremities.len() == 0 { info!("Retained zero extremities when upgrading outlier PDU to timeline PDU with {} previous events, event id: {}", incoming_pdu.prev_events.len(), incoming_pdu.event_id) }
+	
 	debug!(
 		"Retained {} extremities checked against {} prev_events",
 		extremities.len(),

src/service/rooms/timeline/backfill.rs

@@ -1,7 +1,8 @@
 use std::iter::once;
 
+use conduwuit::{Err, PduEvent};
 use conduwuit_core::{
-	Result, debug, debug_warn, implement, info,
+	Result, debug, debug_warn, err, implement, info,
 	matrix::{
 		event::Event,
 		pdu::{PduCount, PduId, RawPduId},
@@ -11,7 +12,7 @@ use conduwuit_core::{
 };
 use futures::{FutureExt, StreamExt};
 use ruma::{
-	RoomId, ServerName,
+	CanonicalJsonObject, EventId, RoomId, ServerName,
 	api::federation,
 	events::{
 		StateEventType, TimelineEventType, room::power_levels::RoomPowerLevelsEventContent,
@@ -100,7 +101,7 @@ pub async fn backfill_if_required(&self, room_id: &RoomId, from: PduCount) -> Re
 		.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}");
 		let response = self
 			.services
 			.sending
@@ -128,10 +129,126 @@ pub async fn backfill_if_required(&self, room_id: &RoomId, from: PduCount) -> Re
 		}
 	}
 
-	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(())
 }
 
+#[implement(super::Service)]
+#[tracing::instrument(name = "get_remote_pdu", level = "debug", skip(self))]
+pub async fn get_remote_pdu(&self, room_id: &RoomId, event_id: &EventId) -> Result<PduEvent> {
+	let local = self.get_pdu(event_id).await;
+	if local.is_ok() {
+		// We already have this PDU, no need to backfill
+		debug!("We already have {event_id} in {room_id}, no need to backfill.");
+		return local;
+	}
+	debug!("Preparing to fetch event {event_id} in room {room_id} from remote servers.");
+	// Similar to backfill_if_required, but only for a single PDU
+	// Fetch a list of servers to try
+	if self
+		.services
+		.state_cache
+		.room_joined_count(room_id)
+		.await
+		.is_ok_and(|count| count <= 1)
+		&& !self
+			.services
+			.state_accessor
+			.is_world_readable(room_id)
+			.await
+	{
+		// Room is empty (1 user or none), there is no one that can backfill
+		return Err!(Request(NotFound("No one can backfill this PDU, room is empty.")));
+	}
+
+	let power_levels: RoomPowerLevelsEventContent = self
+		.services
+		.state_accessor
+		.room_state_get_content(room_id, &StateEventType::RoomPowerLevels, "")
+		.await
+		.unwrap_or_default();
+
+	let room_mods = power_levels.users.iter().filter_map(|(user_id, level)| {
+		if level > &power_levels.users_default && !self.services.globals.user_is_local(user_id) {
+			Some(user_id.server_name())
+		} else {
+			None
+		}
+	});
+
+	let canonical_room_alias_server = once(
+		self.services
+			.state_accessor
+			.get_canonical_alias(room_id)
+			.await,
+	)
+	.filter_map(Result::ok)
+	.map(|alias| alias.server_name().to_owned())
+	.stream();
+	let mut servers = room_mods
+		.stream()
+		.map(ToOwned::to_owned)
+		.chain(canonical_room_alias_server)
+		.chain(
+			self.services
+				.server
+				.config
+				.trusted_servers
+				.iter()
+				.map(ToOwned::to_owned)
+				.stream(),
+		)
+		.ready_filter(|server_name| !self.services.globals.server_is_ours(server_name))
+		.filter_map(|server_name| async move {
+			self.services
+				.state_cache
+				.server_in_room(&server_name, room_id)
+				.await
+				.then_some(server_name)
+		})
+		.boxed();
+
+	while let Some(ref backfill_server) = servers.next().await {
+		info!("Asking {backfill_server} for event {}", event_id);
+		let value = self
+			.services
+			.sending
+			.send_federation_request(backfill_server, federation::event::get_event::v1::Request {
+				event_id: event_id.to_owned(),
+				include_unredacted_content: Some(false),
+			})
+			.await
+			.and_then(|response| {
+				serde_json::from_str::<CanonicalJsonObject>(response.pdu.get()).map_err(|e| {
+					err!(BadServerResponse(debug_warn!(
+						"Error parsing incoming event {e:?} from {backfill_server}"
+					)))
+				})
+			});
+		let pdu = match value {
+			| Ok(value) => {
+				self.services
+					.event_handler
+					.handle_incoming_pdu(backfill_server, &room_id, &event_id, value, false)
+					.boxed()
+					.await?;
+				debug!("Successfully backfilled {event_id} from {backfill_server}");
+				Some(self.get_pdu(event_id).await)
+			},
+			| Err(e) => {
+				warn!("{backfill_server} failed to provide backfill for room {room_id}: {e}");
+				None
+			},
+		};
+		if let Some(pdu) = pdu {
+			debug!("Fetched {event_id} from {backfill_server}");
+			return pdu;
+		}
+	}
+
+	Err!("No servers could be used to fetch {} in {}.", room_id, event_id)
+}
+
 #[implement(super::Service)]
 #[tracing::instrument(skip(self, pdu), level = "debug")]
 pub async fn backfill_pdu(&self, origin: &ServerName, pdu: Box<RawJsonValue>) -> Result<()> {

src/service/rooms/timeline/data.rs

@@ -3,8 +3,7 @@ use std::{borrow::Borrow, sync::Arc};
 use conduwuit::{
 	Err, PduCount, PduEvent, Result, at, err,
 	result::{LogErr, NotFound},
-	utils,
-	utils::stream::TryReadyExt,
+	utils::{self, stream::TryReadyExt},
 };
 use database::{Database, Deserialized, Json, KeyVal, Map};
 use futures::{FutureExt, Stream, TryFutureExt, TryStreamExt, future::select_ok, pin_mut};

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

@@ -10,7 +10,7 @@ use std::{
 
 use base64::{Engine as _, engine::general_purpose::URL_SAFE_NO_PAD};
 use conduwuit_core::{
-	Error, Event, Result, debug, err, error,
+	Error, Event, Result, debug, err, error, info,
 	result::LogErr,
 	trace,
 	utils::{
@@ -142,7 +142,7 @@ impl Service {
 	}
 
 	fn handle_response_err(dest: Destination, statuses: &mut CurTransactionStatus, e: &Error) {
-		debug!(dest = ?dest, "{e:?}");
+		debug!(dest = ?dest, "error response: {e:?}");
 		statuses.entry(dest).and_modify(|e| {
 			*e = match e {
 				| TransactionStatus::Running => TransactionStatus::Failed(1, Instant::now()),
@@ -177,7 +177,21 @@ impl Service {
 		if !new_events.is_empty() {
 			self.db.mark_as_active(new_events.iter());
 
-			let new_events_vec = new_events.into_iter().map(|(_, event)| event).collect();
+			let new_events_vec: Vec<SendingEvent> =
+				new_events.into_iter().map(|(_, event)| event).collect();
+
+			if let Some(status) = statuses.get(&dest.clone()) {
+				if matches!(status, TransactionStatus::Running) {
+					// If the server is in backoff, clear it
+					info!(
+						?dest,
+						"Catching up previously failed destination with {}+ new events",
+						new_events_vec.len()
+					);
+					statuses.insert(dest.clone(), TransactionStatus::Running);
+				}
+			}
+
 			futures.push(self.send_events(dest.clone(), new_events_vec));
 		} else {
 			statuses.remove(dest);
@@ -859,16 +873,24 @@ impl Service {
 			pdus,
 			edus,
 		};
+		let pdu_count = request.pdus.len();
+		let edu_count = request.edus.len();
 
 		let result = self
 			.services
 			.federation
 			.execute_on(&self.services.client.sender, &server, request)
-			.await;
+			.await
+			.inspect(|_| {
+				trace!(%txn_id, %server, "Sent {} PDUs, {} EDUs", pdu_count, edu_count);
+			})
+			.inspect_err(|e| {
+				info!(%txn_id, %server, "Failed to send transaction ({} PDUs, {} EDUs): {e:?}", pdu_count, edu_count);
+			});
 
 		for (event_id, result) in result.iter().flat_map(|resp| resp.pdus.iter()) {
 			if let Err(e) = result {
-				warn!(
+				trace!(
 					%txn_id, %server,
 					"error sending PDU {event_id} to remote server: {e:?}"
 				);