Merge pull request 'Some improvements to Garage internals' (#451) from internals-rework into main

Reviewed-on: https://git.deuxfleurs.fr/Deuxfleurs/garage/pulls/451

8 files changed, 218 insertions, 72 deletions

diff --git a/src/table/Cargo.toml b/src/table/Cargo.toml
index 861e3843..e1a74553 100644
--- a/src/table/Cargo.toml
+++ b/src/table/Cargo.toml
@@ -21,6 +21,7 @@ garage_util = { version = "0.8.1", path = "../util" }
 opentelemetry = "0.17"
 
 async-trait = "0.1.7"
+arc-swap = "1.0"
 bytes = "1.0"
 hex = "0.4"
 hexdump = "0.1"
diff --git a/src/table/data.rs b/src/table/data.rs
index 93da2110..40856b02 100644
--- a/src/table/data.rs
+++ b/src/table/data.rs
@@ -31,6 +31,10 @@ pub struct TableData<F: TableSchema, R: TableReplication> {
 	pub(crate) merkle_tree: db::Tree,
 	pub(crate) merkle_todo: db::Tree,
 	pub(crate) merkle_todo_notify: Notify,
+
+	pub(crate) insert_queue: db::Tree,
+	pub(crate) insert_queue_notify: Notify,
+
 	pub(crate) gc_todo: CountedTree,
 
 	pub(crate) metrics: TableMetrics,
@@ -53,9 +57,13 @@ where
 			.open_tree(&format!("{}:merkle_todo", F::TABLE_NAME))
 			.expect("Unable to open DB Merkle TODO tree");
 
+		let insert_queue = db
+			.open_tree(&format!("{}:insert_queue", F::TABLE_NAME))
+			.expect("Unable to open insert queue DB tree");
+
 		let gc_todo = db
 			.open_tree(&format!("{}:gc_todo_v2", F::TABLE_NAME))
-			.expect("Unable to open DB tree");
+			.expect("Unable to open GC DB tree");
 		let gc_todo = CountedTree::new(gc_todo).expect("Cannot count gc_todo_v2");
 
 		let metrics = TableMetrics::new(
@@ -74,6 +82,8 @@ where
 			merkle_tree,
 			merkle_todo,
 			merkle_todo_notify: Notify::new(),
+			insert_queue,
+			insert_queue_notify: Notify::new(),
 			gc_todo,
 			metrics,
 		})
@@ -173,9 +183,8 @@ where
 
 	pub(crate) fn update_entry(&self, update_bytes: &[u8]) -> Result<(), Error> {
 		let update = self.decode_entry(update_bytes)?;
-		let tree_key = self.tree_key(update.partition_key(), update.sort_key());
 
-		self.update_entry_with(&tree_key[..], |ent| match ent {
+		self.update_entry_with(update.partition_key(), update.sort_key(), |ent| match ent {
 			Some(mut ent) => {
 				ent.merge(&update);
 				ent
@@ -187,11 +196,14 @@ where
 
 	pub fn update_entry_with(
 		&self,
-		tree_key: &[u8],
+		partition_key: &F::P,
+		sort_key: &F::S,
 		f: impl Fn(Option<F::E>) -> F::E,
 	) -> Result<Option<F::E>, Error> {
+		let tree_key = self.tree_key(partition_key, sort_key);
+
 		let changed = self.store.db().transaction(|mut tx| {
-			let (old_entry, old_bytes, new_entry) = match tx.get(&self.store, tree_key)? {
+			let (old_entry, old_bytes, new_entry) = match tx.get(&self.store, &tree_key)? {
 				Some(old_bytes) => {
 					let old_entry = self.decode_entry(&old_bytes).map_err(db::TxError::Abort)?;
 					let new_entry = f(Some(old_entry.clone()));
@@ -200,23 +212,23 @@ where
 				None => (None, None, f(None)),
 			};
 
-			// Scenario 1: the value changed, so of course there is a change
-			let value_changed = Some(&new_entry) != old_entry.as_ref();
-
+			// Changed can be true in two scenarios
+			// Scenario 1: the actual represented value changed,
+			//   so of course the messagepack encoding changed as well
 			// Scenario 2: the value didn't change but due to a migration in the
-			// data format, the messagepack encoding changed. In this case
-			// we have to write the migrated value in the table and update
-			// the associated Merkle tree entry.
+			//   data format, the messagepack encoding changed. In this case,
+			//   we also have to write the migrated value in the table and update
+			//   the associated Merkle tree entry.
 			let new_bytes = rmp_to_vec_all_named(&new_entry)
 				.map_err(Error::RmpEncode)
 				.map_err(db::TxError::Abort)?;
-			let encoding_changed = Some(&new_bytes[..]) != old_bytes.as_ref().map(|x| &x[..]);
+			let changed = Some(&new_bytes[..]) != old_bytes.as_deref();
 			drop(old_bytes);
 
-			if value_changed || encoding_changed {
-				let new_bytes_hash = blake2sum(&new_bytes[..]);
-				tx.insert(&self.merkle_todo, tree_key, new_bytes_hash.as_slice())?;
-				tx.insert(&self.store, tree_key, new_bytes)?;
+			if changed {
+				let new_bytes_hash = blake2sum(&new_bytes);
+				tx.insert(&self.merkle_todo, &tree_key, new_bytes_hash.as_slice())?;
+				tx.insert(&self.store, &tree_key, new_bytes)?;
 
 				self.instance
 					.updated(&mut tx, old_entry.as_ref(), Some(&new_entry))?;
@@ -242,7 +254,7 @@ where
 				let pk_hash = Hash::try_from(&tree_key[..32]).unwrap();
 				let nodes = self.replication.write_nodes(&pk_hash);
 				if nodes.first() == Some(&self.system.id) {
-					GcTodoEntry::new(tree_key.to_vec(), new_bytes_hash).save(&self.gc_todo)?;
+					GcTodoEntry::new(tree_key, new_bytes_hash).save(&self.gc_todo)?;
 				}
 			}
 
@@ -258,10 +270,11 @@ where
 			.db()
 			.transaction(|mut tx| match tx.get(&self.store, k)? {
 				Some(cur_v) if cur_v == v => {
+					let old_entry = self.decode_entry(v).map_err(db::TxError::Abort)?;
+
 					tx.remove(&self.store, k)?;
 					tx.insert(&self.merkle_todo, k, vec![])?;
 
-					let old_entry = self.decode_entry(v).map_err(db::TxError::Abort)?;
 					self.instance.updated(&mut tx, Some(&old_entry), None)?;
 					Ok(true)
 				}
@@ -285,10 +298,11 @@ where
 			.db()
 			.transaction(|mut tx| match tx.get(&self.store, k)? {
 				Some(cur_v) if blake2sum(&cur_v[..]) == vhash => {
+					let old_entry = self.decode_entry(&cur_v[..]).map_err(db::TxError::Abort)?;
+
 					tx.remove(&self.store, k)?;
 					tx.insert(&self.merkle_todo, k, vec![])?;
 
-					let old_entry = self.decode_entry(&cur_v[..]).map_err(db::TxError::Abort)?;
 					self.instance.updated(&mut tx, Some(&old_entry), None)?;
 					Ok(true)
 				}
@@ -302,6 +316,32 @@ where
 		Ok(removed)
 	}
 
+	// ---- Insert queue functions ----
+
+	pub(crate) fn queue_insert(
+		&self,
+		tx: &mut db::Transaction,
+		ins: &F::E,
+	) -> db::TxResult<(), Error> {
+		let tree_key = self.tree_key(ins.partition_key(), ins.sort_key());
+
+		let new_entry = match tx.get(&self.insert_queue, &tree_key)? {
+			Some(old_v) => {
+				let mut entry = self.decode_entry(&old_v).map_err(db::TxError::Abort)?;
+				entry.merge(ins);
+				rmp_to_vec_all_named(&entry)
+			}
+			None => rmp_to_vec_all_named(ins),
+		};
+		let new_entry = new_entry
+			.map_err(Error::RmpEncode)
+			.map_err(db::TxError::Abort)?;
+		tx.insert(&self.insert_queue, &tree_key, new_entry)?;
+		self.insert_queue_notify.notify_one();
+
+		Ok(())
+	}
+
 	// ---- Utility functions ----
 
 	pub fn tree_key(&self, p: &F::P, s: &F::S) -> Vec<u8> {
diff --git a/src/table/gc.rs b/src/table/gc.rs
index cfdc9d2d..90594fba 100644
--- a/src/table/gc.rs
+++ b/src/table/gc.rs
@@ -54,24 +54,25 @@ where
 	F: TableSchema + 'static,
 	R: TableReplication + 'static,
 {
-	pub(crate) fn launch(system: Arc<System>, data: Arc<TableData<F, R>>) -> Arc<Self> {
+	pub(crate) fn new(system: Arc<System>, data: Arc<TableData<F, R>>) -> Arc<Self> {
 		let endpoint = system
 			.netapp
 			.endpoint(format!("garage_table/gc.rs/Rpc:{}", F::TABLE_NAME));
 
 		let gc = Arc::new(Self {
-			system: system.clone(),
+			system,
 			data,
 			endpoint,
 		});
-
 		gc.endpoint.set_handler(gc.clone());
 
-		system.background.spawn_worker(GcWorker::new(gc.clone()));
-
 		gc
 	}
 
+	pub(crate) fn spawn_workers(self: &Arc<Self>, bg: &BackgroundRunner) {
+		bg.spawn_worker(GcWorker::new(self.clone()));
+	}
+
 	async fn gc_loop_iter(&self) -> Result<Option<Duration>, Error> {
 		let now = now_msec();
 
@@ -347,10 +348,7 @@ where
 		}
 	}
 
-	async fn wait_for_work(&mut self, must_exit: &watch::Receiver<bool>) -> WorkerState {
-		if *must_exit.borrow() {
-			return WorkerState::Done;
-		}
+	async fn wait_for_work(&mut self) -> WorkerState {
 		tokio::time::sleep(self.wait_delay).await;
 		WorkerState::Busy
 	}
diff --git a/src/table/lib.rs b/src/table/lib.rs
index b0153e9a..fdf114a6 100644
--- a/src/table/lib.rs
+++ b/src/table/lib.rs
@@ -4,16 +4,18 @@
 #[macro_use]
 extern crate tracing;
 
-mod metrics;
 pub mod schema;
 pub mod util;
 
 pub mod data;
+pub mod replication;
+pub mod table;
+
 mod gc;
 mod merkle;
-pub mod replication;
+mod metrics;
+mod queue;
 mod sync;
-pub mod table;
 
 pub use schema::*;
 pub use table::*;
diff --git a/src/table/merkle.rs b/src/table/merkle.rs
index e977bfb5..736354fa 100644
--- a/src/table/merkle.rs
+++ b/src/table/merkle.rs
@@ -3,6 +3,7 @@ use std::time::Duration;
 
 use async_trait::async_trait;
 use serde::{Deserialize, Serialize};
+use tokio::select;
 use tokio::sync::watch;
 
 use garage_db as db;
@@ -69,17 +70,17 @@ where
 	F: TableSchema + 'static,
 	R: TableReplication + 'static,
 {
-	pub(crate) fn launch(background: &BackgroundRunner, data: Arc<TableData<F, R>>) -> Arc<Self> {
+	pub(crate) fn new(data: Arc<TableData<F, R>>) -> Arc<Self> {
 		let empty_node_hash = blake2sum(&rmp_to_vec_all_named(&MerkleNode::Empty).unwrap()[..]);
 
-		let ret = Arc::new(Self {
+		Arc::new(Self {
 			data,
 			empty_node_hash,
-		});
-
-		background.spawn_worker(MerkleWorker(ret.clone()));
+		})
+	}
 
-		ret
+	pub(crate) fn spawn_workers(self: &Arc<Self>, background: &BackgroundRunner) {
+		background.spawn_worker(MerkleWorker(self.clone()));
 	}
 
 	fn updater_loop_iter(&self) -> Result<WorkerState, Error> {
@@ -339,11 +340,11 @@ where
 		.unwrap()
 	}
 
-	async fn wait_for_work(&mut self, must_exit: &watch::Receiver<bool>) -> WorkerState {
-		if *must_exit.borrow() {
-			return WorkerState::Done;
+	async fn wait_for_work(&mut self) -> WorkerState {
+		select! {
+			_ = tokio::time::sleep(Duration::from_secs(60)) => (),
+			_ = self.0.data.merkle_todo_notify.notified() => (),
 		}
-		tokio::time::sleep(Duration::from_secs(10)).await;
 		WorkerState::Busy
 	}
 }
diff --git a/src/table/queue.rs b/src/table/queue.rs
new file mode 100644
index 00000000..860f20d3
--- /dev/null
+++ b/src/table/queue.rs
@@ -0,0 +1,81 @@
+use std::sync::Arc;
+use std::time::Duration;
+
+use async_trait::async_trait;
+use tokio::select;
+use tokio::sync::watch;
+
+use garage_util::background::*;
+use garage_util::error::Error;
+
+use crate::replication::*;
+use crate::schema::*;
+use crate::table::*;
+
+const BATCH_SIZE: usize = 100;
+
+pub(crate) struct InsertQueueWorker<F, R>(pub(crate) Arc<Table<F, R>>)
+where
+	F: TableSchema + 'static,
+	R: TableReplication + 'static;
+
+#[async_trait]
+impl<F, R> Worker for InsertQueueWorker<F, R>
+where
+	F: TableSchema + 'static,
+	R: TableReplication + 'static,
+{
+	fn name(&self) -> String {
+		format!("{} queue", F::TABLE_NAME)
+	}
+
+	fn status(&self) -> WorkerStatus {
+		WorkerStatus {
+			queue_length: Some(self.0.data.insert_queue.len().unwrap_or(0) as u64),
+			..Default::default()
+		}
+	}
+
+	async fn work(&mut self, _must_exit: &mut watch::Receiver<bool>) -> Result<WorkerState, Error> {
+		let mut kv_pairs = vec![];
+		let mut values = vec![];
+
+		for entry_kv in self.0.data.insert_queue.iter()? {
+			let (k, v) = entry_kv?;
+
+			values.push(self.0.data.decode_entry(&v)?);
+			kv_pairs.push((k, v));
+
+			if kv_pairs.len() > BATCH_SIZE {
+				break;
+			}
+		}
+
+		if kv_pairs.is_empty() {
+			return Ok(WorkerState::Idle);
+		}
+
+		self.0.insert_many(values).await?;
+
+		self.0.data.insert_queue.db().transaction(|mut tx| {
+			for (k, v) in kv_pairs.iter() {
+				if let Some(v2) = tx.get(&self.0.data.insert_queue, k)? {
+					if &v2 == v {
+						tx.remove(&self.0.data.insert_queue, k)?;
+					}
+				}
+			}
+			Ok(())
+		})?;
+
+		Ok(WorkerState::Busy)
+	}
+
+	async fn wait_for_work(&mut self) -> WorkerState {
+		select! {
+			_ = tokio::time::sleep(Duration::from_secs(600)) => (),
+			_ = self.0.data.insert_queue_notify.notified() => (),
+		}
+		WorkerState::Busy
+	}
+}
diff --git a/src/table/sync.rs b/src/table/sync.rs
index af7aa640..d6d272ab 100644
--- a/src/table/sync.rs
+++ b/src/table/sync.rs
@@ -2,6 +2,7 @@ use std::collections::VecDeque;
 use std::sync::Arc;
 use std::time::{Duration, Instant};
 
+use arc_swap::ArcSwapOption;
 use async_trait::async_trait;
 use futures_util::stream::*;
 use opentelemetry::KeyValue;
@@ -13,7 +14,7 @@ use tokio::sync::{mpsc, watch};
 
 use garage_util::background::*;
 use garage_util::data::*;
-use garage_util::error::Error;
+use garage_util::error::{Error, OkOrMessage};
 
 use garage_rpc::ring::*;
 use garage_rpc::system::System;
@@ -32,7 +33,7 @@ pub struct TableSyncer<F: TableSchema + 'static, R: TableReplication + 'static>
 	data: Arc<TableData<F, R>>,
 	merkle: Arc<MerkleUpdater<F, R>>,
 
-	add_full_sync_tx: mpsc::UnboundedSender<()>,
+	add_full_sync_tx: ArcSwapOption<mpsc::UnboundedSender<()>>,
 	endpoint: Arc<Endpoint<SyncRpc, Self>>,
 }
 
@@ -65,7 +66,7 @@ where
 	F: TableSchema + 'static,
 	R: TableReplication + 'static,
 {
-	pub(crate) fn launch(
+	pub(crate) fn new(
 		system: Arc<System>,
 		data: Arc<TableData<F, R>>,
 		merkle: Arc<MerkleUpdater<F, R>>,
@@ -74,34 +75,40 @@ where
 			.netapp
 			.endpoint(format!("garage_table/sync.rs/Rpc:{}", F::TABLE_NAME));
 
-		let (add_full_sync_tx, add_full_sync_rx) = mpsc::unbounded_channel();
-
 		let syncer = Arc::new(Self {
-			system: system.clone(),
+			system,
 			data,
 			merkle,
-			add_full_sync_tx,
+			add_full_sync_tx: ArcSwapOption::new(None),
 			endpoint,
 		});
-
 		syncer.endpoint.set_handler(syncer.clone());
 
-		system.background.spawn_worker(SyncWorker {
-			syncer: syncer.clone(),
-			ring_recv: system.ring.clone(),
-			ring: system.ring.borrow().clone(),
+		syncer
+	}
+
+	pub(crate) fn spawn_workers(self: &Arc<Self>, bg: &BackgroundRunner) {
+		let (add_full_sync_tx, add_full_sync_rx) = mpsc::unbounded_channel();
+		self.add_full_sync_tx
+			.store(Some(Arc::new(add_full_sync_tx)));
+
+		bg.spawn_worker(SyncWorker {
+			syncer: self.clone(),
+			ring_recv: self.system.ring.clone(),
+			ring: self.system.ring.borrow().clone(),
 			add_full_sync_rx,
 			todo: vec![],
 			next_full_sync: Instant::now() + Duration::from_secs(20),
 		});
-
-		syncer
 	}
 
-	pub fn add_full_sync(&self) {
-		if self.add_full_sync_tx.send(()).is_err() {
-			error!("({}) Could not add full sync", F::TABLE_NAME);
-		}
+	pub fn add_full_sync(&self) -> Result<(), Error> {
+		let tx = self.add_full_sync_tx.load();
+		let tx = tx
+			.as_ref()
+			.ok_or_message("table sync worker is not running")?;
+		tx.send(()).ok_or_message("send error")?;
+		Ok(())
 	}
 
 	// ----
@@ -586,10 +593,7 @@ impl<F: TableSchema + 'static, R: TableReplication + 'static> Worker for SyncWor
 		}
 	}
 
-	async fn wait_for_work(&mut self, must_exit: &watch::Receiver<bool>) -> WorkerState {
-		if *must_exit.borrow() {
-			return WorkerState::Done;
-		}
+	async fn wait_for_work(&mut self) -> WorkerState {
 		select! {
 			s = self.add_full_sync_rx.recv() => {
 				if let Some(()) = s {
diff --git a/src/table/table.rs b/src/table/table.rs
index 8a66c420..bbcd5971 100644
--- a/src/table/table.rs
+++ b/src/table/table.rs
@@ -14,6 +14,7 @@ use opentelemetry::{
 
 use garage_db as db;
 
+use garage_util::background::BackgroundRunner;
 use garage_util::data::*;
 use garage_util::error::Error;
 use garage_util::metrics::RecordDuration;
@@ -25,6 +26,7 @@ use crate::crdt::Crdt;
 use crate::data::*;
 use crate::gc::*;
 use crate::merkle::*;
+use crate::queue::InsertQueueWorker;
 use crate::replication::*;
 use crate::schema::*;
 use crate::sync::*;
@@ -35,6 +37,7 @@ pub struct Table<F: TableSchema + 'static, R: TableReplication + 'static> {
 	pub data: Arc<TableData<F, R>>,
 	pub merkle_updater: Arc<MerkleUpdater<F, R>>,
 	pub syncer: Arc<TableSyncer<F, R>>,
+	gc: Arc<TableGc<F, R>>,
 	endpoint: Arc<Endpoint<TableRpc<F>, Self>>,
 }
 
@@ -75,15 +78,16 @@ where
 
 		let data = TableData::new(system.clone(), instance, replication, db);
 
-		let merkle_updater = MerkleUpdater::launch(&system.background, data.clone());
+		let merkle_updater = MerkleUpdater::new(data.clone());
 
-		let syncer = TableSyncer::launch(system.clone(), data.clone(), merkle_updater.clone());
-		TableGc::launch(system.clone(), data.clone());
+		let syncer = TableSyncer::new(system.clone(), data.clone(), merkle_updater.clone());
+		let gc = TableGc::new(system.clone(), data.clone());
 
 		let table = Arc::new(Self {
 			system,
 			data,
 			merkle_updater,
+			gc,
 			syncer,
 			endpoint,
 		});
@@ -93,6 +97,13 @@ where
 		table
 	}
 
+	pub fn spawn_workers(self: &Arc<Self>, bg: &BackgroundRunner) {
+		self.merkle_updater.spawn_workers(bg);
+		self.syncer.spawn_workers(bg);
+		self.gc.spawn_workers(bg);
+		bg.spawn_worker(InsertQueueWorker(self.clone()));
+	}
+
 	pub async fn insert(&self, e: &F::E) -> Result<(), Error> {
 		let tracer = opentelemetry::global::tracer("garage_table");
 		let span = tracer.start(format!("{} insert", F::TABLE_NAME));
@@ -128,6 +139,11 @@ where
 		Ok(())
 	}
 
+	/// Insert item locally
+	pub fn queue_insert(&self, tx: &mut db::Transaction, e: &F::E) -> db::TxResult<(), Error> {
+		self.data.queue_insert(tx, e)
+	}
+
 	pub async fn insert_many<I, IE>(&self, entries: I) -> Result<(), Error>
 	where
 		I: IntoIterator<Item = IE> + Send + Sync,
@@ -259,9 +275,11 @@ where
 			if not_all_same {
 				let self2 = self.clone();
 				let ent2 = ret_entry.clone();
-				self.system
-					.background
-					.spawn_cancellable(async move { self2.repair_on_read(&who[..], ent2).await });
+				tokio::spawn(async move {
+					if let Err(e) = self2.repair_on_read(&who[..], ent2).await {
+						warn!("Error doing repair on read: {}", e);
+					}
+				});
 			}
 		}
 
@@ -358,11 +376,12 @@ where
 				.into_iter()
 				.map(|k| ret.get(&k).unwrap().clone())
 				.collect::<Vec<_>>();
-			self.system.background.spawn_cancellable(async move {
+			tokio::spawn(async move {
 				for v in to_repair {
-					self2.repair_on_read(&who[..], v).await?;
+					if let Err(e) = self2.repair_on_read(&who[..], v).await {
+						warn!("Error doing repair on read: {}", e);
+					}
 				}
-				Ok(())
 			});
 		}