fix(job-run): remove unneded retry_bail!

svc/pkg/cluster/worker/src/workers/nomad_node_registered.rs

@@ -38,7 +38,7 @@ async fn worker(
 
 	// Insert metrics
 	if let Some(install_complete_ts) = install_complete_ts {
-	insert_metrics(ctx, datacenter_id, nomad_join_ts, install_complete_ts).await?;
+		insert_metrics(ctx, datacenter_id, nomad_join_ts, install_complete_ts).await?;
 	} else {
 		tracing::warn!("missing install_complete_ts for nomad-node-registered");
 	}

svc/pkg/job-run/worker/src/workers/cleanup.rs

@@ -27,12 +27,16 @@ async fn worker(ctx: &OperationContext<job_run::msg::cleanup::Message>) -> Globa
 		})
 		.await?
 	else {
-		if ctx.req_dt() > util::duration::minutes(5) {
-			tracing::error!("discarding stale message");
-			return Ok(());
-		} else {
-			retry_bail!("run not found, may be race condition with insertion");
-		}
+		// if ctx.req_dt() > util::duration::minutes(5) {
+		// 	tracing::error!("discarding stale message");
+		// 	return Ok(());
+		// } else {
+		// 	retry_bail!("run not found, may be race condition with insertion");
+		// }
+
+		// TODO: This has amplifying failures, so we just fail once here
+		tracing::error!("job run not found, may have leaked");
+		return Ok(());
 	};
 
 	tracing::info!("removing from cache");

svc/pkg/mm/standalone/gc/src/lib.rs

@@ -1,6 +1,7 @@
 use proto::backend::pkg::*;
 use redis::AsyncCommands;
 use rivet_operation::prelude::*;
+use rivet_pools::prelude::redis::Commands;
 
 #[tracing::instrument(skip_all)]
 pub async fn run_from_env(ts: i64, pools: rivet_pools::Pools) -> GlobalResult<()> {
@@ -86,11 +87,13 @@ async fn cull_unregistered_players(
 	mut redis_mm: RedisPool,
 	client: chirp_client::Client,
 ) -> GlobalResult<()> {
-	// We don't remove from the set here since this will be removed in the mm-player-remove
-	// service.
-	let remove_player_ids = redis_mm
-		.zrangebyscore::<_, _, _, Vec<String>>(util_mm::key::player_unregistered(), 0, ts as isize)
+	let remove_player_ids = redis::pipe()
+		.zrangebyscore(util_mm::key::player_unregistered(), 0, ts as isize)
+		.zrembyscore(util_mm::key::player_unregistered(), 0, ts as isize)
+		.ignore()
+		.query_async::<_, (Vec<String>,)>(&mut redis_mm)
 		.await?
+		.0
 		.into_iter()
 		.filter_map(|x| util::uuid::parse(&x).ok())
 		.collect::<Vec<_>>();
@@ -115,11 +118,13 @@ async fn cull_auto_remove_players(
 	mut redis_mm: RedisPool,
 	client: chirp_client::Client,
 ) -> GlobalResult<()> {
-	// We don't remove from the set here since this will be removed in the mm-player-remove
-	// service.
-	let remove_player_ids = redis_mm
-		.zrangebyscore::<_, _, _, Vec<String>>(util_mm::key::player_auto_remove(), 0, ts as isize)
+	let remove_player_ids = redis::pipe()
+		.zrangebyscore(util_mm::key::player_auto_remove(), 0, ts as isize)
+		.zrembyscore(util_mm::key::player_auto_remove(), 0, ts as isize)
+		.ignore()
+		.query_async::<_, (Vec<String>,)>(&mut redis_mm)
 		.await?
+		.0
 		.into_iter()
 		.filter_map(|x| util::uuid::parse(&x).ok())
 		.collect::<Vec<_>>();

svc/pkg/mm/worker/src/workers/lobby_cleanup.rs

@@ -95,7 +95,11 @@ async fn worker(ctx: &OperationContext<mm::msg::lobby_cleanup::Message>) -> Glob
 			tracing::error!("discarding stale message");
 			return Ok(());
 		} else {
-			retry_bail!("lobby not found, may be race condition with insertion");
+			// retry_bail!("lobby not found, may be race condition with insertion");
+
+			// TODO: This has amplifying failures, so we just fail once here
+			tracing::error!("lobby not found, may have leaked");
+			return Ok(());
 		}
 	};
 

svc/pkg/mm/worker/src/workers/lobby_stop.rs

@@ -43,7 +43,11 @@ async fn worker(ctx: &OperationContext<mm::msg::lobby_stop::Message>) -> GlobalR
 			tracing::error!("discarding stale message");
 			return Ok(());
 		} else {
-			retry_bail!("lobby not found, may be race condition with insertion");
+			// retry_bail!("lobby not found, may be race condition with insertion");
+
+			// TODO: This has amplifying failures, so we just fail once here
+			tracing::error!("lobby not found, may have leaked");
+			return Ok(());
 		}
 	};
 

svc/pkg/mm/worker/src/workers/player_remove.rs

@@ -83,7 +83,11 @@ async fn worker(ctx: &OperationContext<mm::msg::player_remove::Message>) -> Glob
 		tracing::error!("discarding stale message");
 		return Ok(());
 	} else {
-		retry_bail!("player not found, may be race condition with insertion");
+		// retry_bail!("player not found, may be race condition with insertion");
+
+		// TODO: This has amplifying failures, so we just fail once here
+		tracing::error!("player not found, may have leaked");
+		return Ok(());
 	};
 
 	// Validate lobby