Take replication lag into account while selecting primary candidate (v…

…itessio#14634)

Signed-off-by: Manan Gupta <[email protected]>

changelog/19.0/19.0.0/summary.md

@@ -12,6 +12,8 @@
     - [`FOREIGN_KEY_CHECKS` is now a Vitess Aware Variable](#fk-checks-vitess-aware)
   - **[Query Compatibility](#query-compatibility)**
     - [`SHOW VSCHEMA KEYSPACES` Query](#show-vschema-keyspaces)
+  - **[Planned Reparent Shard](#planned-reparent-shard)**
+    - [`--tolerable-replication-lag` Sub-flag](#tolerable-repl-lag)
 
 ## <a id="major-changes"/>Major Changes
 
@@ -66,3 +68,13 @@ mysql> show vschema keyspaces;
 +----------+---------+-------------+---------+
 2 rows in set (0.01 sec)
 ```
+
+### <a id="planned-reparent-shard"/>Planned Reparent Shard
+
+#### <a id="tolerable-repl-lag"/>`--tolerable-replication-lag` Sub-flag
+
+A new sub-flag `--tolerable-replication-lag` has been added to the command `PlannedReparentShard` that allows users to specify the amount of replication lag that is considered acceptable for a tablet to be eligible for promotion when Vitess makes the choice of a new primary.
+This feature is opt-in and not specifying this sub-flag makes Vitess ignore the replication lag entirely.
+
+A new flag in VTOrc with the same name has been added to control the behaviour of the PlannedReparentShard calls that VTOrc issues.
+

go/cmd/vtctldclient/command/reparents.go

@@ -183,9 +183,10 @@ func commandInitShardPrimary(cmd *cobra.Command, args []string) error {
 }
 
 var plannedReparentShardOptions = struct {
-	NewPrimaryAliasStr   string
-	AvoidPrimaryAliasStr string
-	WaitReplicasTimeout  time.Duration
+	NewPrimaryAliasStr      string
+	AvoidPrimaryAliasStr    string
+	WaitReplicasTimeout     time.Duration
+	TolerableReplicationLag time.Duration
 }{}
 
 func commandPlannedReparentShard(cmd *cobra.Command, args []string) error {
@@ -216,11 +217,12 @@ func commandPlannedReparentShard(cmd *cobra.Command, args []string) error {
 	cli.FinishedParsing(cmd)
 
 	resp, err := client.PlannedReparentShard(commandCtx, &vtctldatapb.PlannedReparentShardRequest{
-		Keyspace:            keyspace,
-		Shard:               shard,
-		NewPrimary:          newPrimaryAlias,
-		AvoidPrimary:        avoidPrimaryAlias,
-		WaitReplicasTimeout: protoutil.DurationToProto(plannedReparentShardOptions.WaitReplicasTimeout),
+		Keyspace:                keyspace,
+		Shard:                   shard,
+		NewPrimary:              newPrimaryAlias,
+		AvoidPrimary:            avoidPrimaryAlias,
+		WaitReplicasTimeout:     protoutil.DurationToProto(plannedReparentShardOptions.WaitReplicasTimeout),
+		TolerableReplicationLag: protoutil.DurationToProto(plannedReparentShardOptions.TolerableReplicationLag),
 	})
 	if err != nil {
 		return err
@@ -292,6 +294,7 @@ func init() {
 	Root.AddCommand(InitShardPrimary)
 
 	PlannedReparentShard.Flags().DurationVar(&plannedReparentShardOptions.WaitReplicasTimeout, "wait-replicas-timeout", topo.RemoteOperationTimeout, "Time to wait for replicas to catch up on replication both before and after reparenting.")
+	PlannedReparentShard.Flags().DurationVar(&plannedReparentShardOptions.TolerableReplicationLag, "tolerable-replication-lag", 0, "Amount of replication lag that is considered acceptable for a tablet to be eligible for promotion when Vitess makes the choice of a new primary.")
 	PlannedReparentShard.Flags().StringVar(&plannedReparentShardOptions.NewPrimaryAliasStr, "new-primary", "", "Alias of a tablet that should be the new primary.")
 	PlannedReparentShard.Flags().StringVar(&plannedReparentShardOptions.AvoidPrimaryAliasStr, "avoid-primary", "", "Alias of a tablet that should not be the primary; i.e. \"reparent to any other tablet if this one is the primary\".")
 	Root.AddCommand(PlannedReparentShard)

go/flags/endtoend/vtorc.txt

@@ -86,6 +86,7 @@ Flags:
       --tablet_manager_grpc_key string                              the key to use to connect
       --tablet_manager_grpc_server_name string                      the server name to use to validate server certificate
       --tablet_manager_protocol string                              Protocol to use to make tabletmanager RPCs to vttablets. (default "grpc")
+      --tolerable-replication-lag duration                          Amount of replication lag that is considered acceptable for a tablet to be eligible for promotion when Vitess makes the choice of a new primary in PRS
       --topo-information-refresh-duration duration                  Timer duration on which VTOrc refreshes the keyspace and vttablet records from the topology server (default 15s)
       --topo_consul_lock_delay duration                             LockDelay for consul session. (default 15s)
       --topo_consul_lock_session_checks string                      List of checks for consul session. (default "serfHealth")