From 2987323351492397820f904820a09cd0591eac88 Mon Sep 17 00:00:00 2001
From: Tanjin Xu <109303790+tanjinx@users.noreply.github.com>
Date: Tue, 9 May 2023 11:01:47 -0700
Subject: [PATCH] `slack-vitess-r14.0.5:`cherrypick: backport rdonly patch
 (#77)

* backport pr 29

* fix missing variable

* fix missing patch

* update help text
---
 go/flags/endtoend/vtgate.txt  |  2 ++
 go/vt/vtgate/tabletgateway.go | 23 +++++++++++++++++++++--
 2 files changed, 23 insertions(+), 2 deletions(-)

diff --git a/go/flags/endtoend/vtgate.txt b/go/flags/endtoend/vtgate.txt
index 4c7f90efd07..7a1c0c853ee 100644
--- a/go/flags/endtoend/vtgate.txt
+++ b/go/flags/endtoend/vtgate.txt
@@ -67,6 +67,8 @@ Usage of vtgate:
 	gate server query cache size, maximum number of queries to be cached. vtgate analyzes every incoming query and generate a query plan, these plans are being cached in a cache. This config controls the expected amount of unique entries in the cache. (default 5000)
   --gateway_initial_tablet_timeout duration
 	At startup, the tabletGateway will wait up to this duration to get at least one tablet per keyspace/shard/tablet type (default 30s)
+  --gateway_route_replica_to_rdonly
+	route REPLICA queries to RDONLY tablets as well as REPLICA tablets
   --grpc_auth_mode string
 	Which auth plugin implementation to use (eg: static)
   --grpc_auth_mtls_allowed_substrings string
diff --git a/go/vt/vtgate/tabletgateway.go b/go/vt/vtgate/tabletgateway.go
index a03b4d02710..17d7018eb12 100644
--- a/go/vt/vtgate/tabletgateway.go
+++ b/go/vt/vtgate/tabletgateway.go
@@ -49,7 +49,8 @@ var (
 	bufferImplementation = flag.String("buffer_implementation", "keyspace_events", "Allowed values: healthcheck (legacy implementation), keyspace_events (default)")
 	initialTabletTimeout = flag.Duration("gateway_initial_tablet_timeout", 30*time.Second, "At startup, the tabletGateway will wait up to this duration to get at least one tablet per keyspace/shard/tablet type")
 	// retryCount is the number of times a query will be retried on error
-	retryCount = flag.Int("retry-count", 2, "retry count")
+	retryCount           = flag.Int("retry-count", 2, "retry count")
+	routeReplicaToRdonly = flag.Bool("gateway_route_replica_to_rdonly", false, "route REPLICA queries to RDONLY tablets as well as REPLICA tablets")
 )
 
 // TabletGateway implements the Gateway interface.
@@ -278,6 +279,20 @@ func (gw *TabletGateway) withRetry(ctx context.Context, target *querypb.Target,
 		}
 
 		tablets := gw.hc.GetHealthyTabletStats(target)
+
+		// temporary hack to enable REPLICA type queries to address both REPLICA tablets and RDONLY tablets
+		// original commit - https://github.com/tinyspeck/vitess/pull/166/commits/2552b4ce25a9fdb41ff07fa69f2ccf485fea83ac
+		// discoverygateway patch - https://github.com/slackhq/vitess/commit/47adb7c8fc720cb4cb7a090530b3e88d310ff6d3
+		if *routeReplicaToRdonly && target.TabletType == topodatapb.TabletType_REPLICA {
+			// Create a new target for the same original keyspace/shard, but RDONLY tablet type.
+			rdonlyTarget := &querypb.Target{
+				Keyspace:   target.Keyspace,
+				Shard:      target.Shard,
+				TabletType: topodatapb.TabletType_RDONLY,
+			}
+			tablets = append(tablets, gw.hc.GetHealthyTabletStats(rdonlyTarget)...)
+		}
+
 		if len(tablets) == 0 {
 			// if we have a keyspace event watcher, check if the reason why our primary is not available is that it's currently being resharded
 			// or if a reparent operation is in progress.
@@ -326,7 +341,11 @@ func (gw *TabletGateway) withRetry(ctx context.Context, target *querypb.Target,
 
 		startTime := time.Now()
 		var canRetry bool
-		canRetry, err = inner(ctx, target, th.Conn)
+		if *routeReplicaToRdonly && target.TabletType == topodatapb.TabletType_REPLICA {
+			canRetry, err = inner(ctx, th.Target, th.Conn)
+		} else {
+			canRetry, err = inner(ctx, target, th.Conn)
+		}
 		gw.updateStats(target, startTime, err)
 		if canRetry {
 			invalidTablets[topoproto.TabletAliasString(tabletLastUsed.Alias)] = true