src/block/manager.rs

@@ -734,12 +734,9 @@ impl Worker for ResyncWorker {
 	) -> Result<WorkerStatus, Error> {
 		self.tranquilizer.reset();
 		match self.manager.resync_iter().await {
-			Ok(ResyncIterResult::BusyDidSomething) => {
-				self.tranquilizer
-					.tranquilize(self.manager.background_tranquility)
-					.await;
-				Ok(WorkerStatus::Busy)
-			}
+			Ok(ResyncIterResult::BusyDidSomething) => Ok(self
+				.tranquilizer
+				.tranquilize_worker(self.manager.background_tranquility)),
 			Ok(ResyncIterResult::BusyDidNothing) => Ok(WorkerStatus::Busy),
 			Ok(ResyncIterResult::IdleFor(delay)) => {
 				self.next_delay = delay;
@@ -750,10 +747,8 @@ impl Worker for ResyncWorker {
 				// We don't really know how to handle them so just ¯\_(ツ)_/¯
 				// (there is kind of an assumption that Sled won't error on us,
 				// if it does there is not much we can do -- TODO should we just panic?)
-				error!(
-					"Could not do a resync iteration: {} (this is a very bad error)",
-					e
-				);
+				// Here we just give the error to the worker manager,
+				// it will print it to the logs and increment a counter
 				Err(e.into())
 			}
 		}

src/block/repair.rs

@@ -138,8 +138,7 @@ impl Worker for ScrubWorker {
 		self.tranquilizer.reset();
 		if let Some(hash) = self.iterator.next().await? {
 			let _ = self.manager.read_block(&hash).await;
-			self.tranquilizer.tranquilize(self.tranquility).await;
-			Ok(WorkerStatus::Busy)
+			Ok(self.tranquilizer.tranquilize_worker(self.tranquility))
 		} else {
 			Ok(WorkerStatus::Done)
 		}

src/garage/admin.rs

@@ -835,7 +835,9 @@ impl AdminRpcHandler {
 				let workers = if busy {
 					workers
 						.into_iter()
-						.filter(|(_, w)| w.status == WorkerStatus::Busy)
+						.filter(|(_, w)| {
+							matches!(w.status, WorkerStatus::Busy | WorkerStatus::Throttled(_))
+						})
 						.collect()
 				} else {
 					workers

src/garage/cli/util.rs

@@ -242,7 +242,7 @@ pub fn print_worker_info(wi: HashMap<usize, WorkerInfo>) {
 	wi.sort_by_key(|(tid, info)| {
 		(
 			match info.status {
-				WorkerStatus::Busy => 0,
+				WorkerStatus::Busy | WorkerStatus::Throttled(_) => 0,
 				WorkerStatus::Idle => 1,
 				WorkerStatus::Done => 2,
 			},
@@ -256,6 +256,20 @@ pub fn print_worker_info(wi: HashMap<usize, WorkerInfo>) {
 		if let Some(i) = &info.info {
 			table.push(format!("\t\t{}", i));
 		}
+		if info.consecutive_errors > 0 {
+			table.push(format!(
+				"\t\t{} CONSECUTIVE ERRORS ({} total), last: {}",
+				info.consecutive_errors,
+				info.errors,
+				info.last_error.as_deref().unwrap_or("(?)")
+			));
+		} else if info.errors > 0 {
+			table.push(format!(
+				"\t\t{} errors, last: {}",
+				info.errors,
+				info.last_error.as_deref().unwrap_or("(?)")
+			));
+		}
 	}
 	format_table(table);
 }

src/model/index_counter.rs

@@ -2,7 +2,6 @@ use core::ops::Bound;
 use std::collections::{hash_map, BTreeMap, HashMap};
 use std::marker::PhantomData;
 use std::sync::Arc;
-use std::time::Duration;
 
 use async_trait::async_trait;
 use serde::{Deserialize, Serialize};
@@ -408,6 +407,10 @@ impl<T: CountedItem> Worker for IndexPropagatorWorker<T> {
 		format!("{} index counter propagator", T::COUNTER_TABLE_NAME)
 	}
 
+	fn info(&self) -> Option<String> {
+		Some(format!("{} items in queue", self.buf.len()))
+	}
+
 	async fn work(&mut self, must_exit: &mut watch::Receiver<bool>) -> Result<WorkerStatus, Error> {
 		// This loop batches updates to counters to be sent all at once.
 		// They are sent once the propagate_rx channel has been emptied (or is closed).
@@ -429,9 +432,10 @@ impl<T: CountedItem> Worker for IndexPropagatorWorker<T> {
 					error!("({}) Could not propagate {} counter values: {}, these counters will not be updated correctly.", T::COUNTER_TABLE_NAME, self.buf.len(), e);
 					return Ok(WorkerStatus::Done);
 				}
-				warn!("({}) Could not propagate {} counter values: {}, retrying in 5 seconds (retry #{})", T::COUNTER_TABLE_NAME, self.buf.len(), e, self.errors);
-				tokio::time::sleep(Duration::from_secs(5)).await;
-				return Ok(WorkerStatus::Busy);
+				// Propagate error up to worker manager, it will log it, increment a counter,
+				// and sleep for a certain delay (with exponential backoff), waiting for
+				// things to go back to normal
+				return Err(e);
 			} else {
 				self.buf.clear();
 				self.errors = 0;

src/util/background/mod.rs

@@ -31,6 +31,9 @@ pub struct WorkerInfo {
 	pub name: String,
 	pub info: Option<String>,
 	pub status: WorkerStatus,
+	pub errors: usize,
+	pub consecutive_errors: usize,
+	pub last_error: Option<String>,
 }
 
 impl BackgroundRunner {

src/util/background/worker.rs

@@ -17,6 +17,7 @@ use crate::error::Error;
 #[derive(PartialEq, Copy, Clone, Debug, Serialize, Deserialize)]
 pub enum WorkerStatus {
 	Busy,
+	Throttled(f32),
 	Idle,
 	Done,
 }
@@ -82,14 +83,17 @@ impl WorkerProcessor {
 						next_task_id += 1;
 						let stop_signal = self.stop_signal.clone();
 						let stop_signal_worker = self.stop_signal.clone();
-						workers.push(async move {
-							let mut worker = WorkerHandler {
+						let mut worker = WorkerHandler {
 								task_id,
 								stop_signal,
 								stop_signal_worker,
 								worker: new_worker,
 								status: WorkerStatus::Busy,
+								errors: 0,
+								consecutive_errors: 0,
+								last_error: None,
 							};
+						workers.push(async move {
 							worker.step().await;
 							worker
 						}.boxed());
@@ -98,21 +102,31 @@ impl WorkerProcessor {
 				worker = await_next_worker => {
 					if let Some(mut worker) = worker {
 						trace!("{} (TID {}): {:?}", worker.worker.name(), worker.task_id, worker.status);
+
+						// Save worker info
 						let mut wi = self.worker_info.lock().unwrap();
 						match wi.get_mut(&worker.task_id) {
 							Some(i) => {
 								i.status = worker.status;
 								i.info = worker.worker.info();
+								i.errors = worker.errors;
+								i.consecutive_errors = worker.consecutive_errors;
+								if worker.last_error.is_some() {
+									i.last_error = worker.last_error.take();
+								}
 							}
 							None => {
 								wi.insert(worker.task_id, WorkerInfo {
 									name: worker.worker.name(),
 									status: worker.status,
 									info: worker.worker.info(),
+									errors: worker.errors,
+									consecutive_errors: worker.consecutive_errors,
+									last_error: worker.last_error.take(),
 								});
 							}
 						}
-						// TODO save new worker status somewhere
+
 						if worker.status == WorkerStatus::Done {
 							info!("Worker {} (TID {}) exited", worker.worker.name(), worker.task_id);
 						} else {
@@ -169,6 +183,9 @@ struct WorkerHandler {
 	stop_signal_worker: watch::Receiver<bool>,
 	worker: Box<dyn Worker>,
 	status: WorkerStatus,
+	errors: usize,
+	consecutive_errors: usize,
+	last_error: Option<String>,
 }
 
 impl WorkerHandler {
@@ -177,6 +194,7 @@ impl WorkerHandler {
 			WorkerStatus::Busy => match self.worker.work(&mut self.stop_signal).await {
 				Ok(s) => {
 					self.status = s;
+					self.consecutive_errors = 0;
 				}
 				Err(e) => {
 					error!(
@@ -185,11 +203,21 @@ impl WorkerHandler {
 						self.task_id,
 						e
 					);
-					// Sleep a bit so that error won't repeat immediately
-					// (TODO good way to handle errors)
-					tokio::time::sleep(Duration::from_secs(10)).await;
+					self.errors += 1;
+					self.consecutive_errors += 1;
+					self.last_error = Some(format!("{}", e));
+					// Sleep a bit so that error won't repeat immediately, exponential backoff
+					// strategy (min 1sec, max ~60sec)
+					self.status = WorkerStatus::Throttled(
+						(1.5f32).powf(std::cmp::min(10, self.consecutive_errors - 1) as f32),
+					);
 				}
 			},
+			WorkerStatus::Throttled(delay) => {
+				// Sleep for given delay and go back to busy state
+				tokio::time::sleep(Duration::from_secs_f32(delay)).await;
+				self.status = WorkerStatus::Busy;
+			}
 			WorkerStatus::Idle => {
 				if *self.stop_signal.borrow() {
 					select! {

src/util/tranquilizer.rs

@@ -3,6 +3,8 @@ use std::time::{Duration, Instant};
 
 use tokio::time::sleep;
 
+use crate::background::WorkerStatus;
+
 /// A tranquilizer is a helper object that is used to make
 /// background operations not take up too much time.
 ///
@@ -33,7 +35,7 @@ impl Tranquilizer {
 		}
 	}
 
-	pub async fn tranquilize(&mut self, tranquility: u32) {
+	fn tranquilize_internal(&mut self, tranquility: u32) -> Option<Duration> {
 		let observation = Instant::now() - self.last_step_begin;
 
 		self.observations.push_back(observation);
@@ -45,10 +47,25 @@ impl Tranquilizer {
 
 		if !self.observations.is_empty() {
 			let delay = (tranquility * self.sum_observations) / (self.observations.len() as u32);
-			sleep(delay).await;
+			Some(delay)
+		} else {
+			None
 		}
+	}
 
-		self.reset();
+	pub async fn tranquilize(&mut self, tranquility: u32) {
+		if let Some(delay) = self.tranquilize_internal(tranquility) {
+			sleep(delay).await;
+			self.reset();
+		}
+	}
+
+	#[must_use]
+	pub fn tranquilize_worker(&mut self, tranquility: u32) -> WorkerStatus {
+		match self.tranquilize_internal(tranquility) {
+			Some(delay) => WorkerStatus::Throttled(delay.as_secs_f32()),
+			None => WorkerStatus::Busy,
+		}
 	}
 
 	pub fn reset(&mut self) {