Merge cockroachdb#106131

106131: pgrepl: parse replication statements + handle IDENTIFY_SYSTEM r=cucaroach a=otan

Informs: cockroachdb#105130

## pgwire: ban extended protocol with REPLICATION protocol

In line with PostgreSQL, we prohibit the use of the extended protocol
when the replication protocol is in effect.

Release note: None

## pgrepl: implement IDENTIFY_SYSTEM command

This commit implements IDENTIFY_SYSTEM in the replication protocol,
which is used to retrieve metadata about replication state.

Most of this commit involves changing assumptions throughout the
codebase there is 1 parser - there is 2 if replication mode is enabled.
We've also had to change the parser to return `statements.Statement`
instead of `pgrepltree.ReplicationStatement` for compatibility with
`parser.Parse...`.

I've left the `LSN` fields as placeholders for now as we finalise on
what we'll eventually call it.

Release note: None

## sql: no-op SHOW SYNTAX in replication mode

Currently on the CLI, the command must parse on the server for it to
run. For replication related commands, we have not implemented `SHOW
SYNTAX` yet. This is a bit of effort and seems outside the scope of what
we want for now.

For now, handle replication commands by no-oping them.

Release note: None



Co-authored-by: Oliver Tan <[email protected]>

Author: craig[bot]

pkg/BUILD.bazel

@@ -1902,6 +1902,7 @@ GO_TARGETS = [
     "//pkg/sql/parser:parser",
     "//pkg/sql/parser:parser_test",
     "//pkg/sql/pgrepl/lsn:lsn",
+    "//pkg/sql/pgrepl/lsnutil:lsnutil",
     "//pkg/sql/pgrepl/pgreplparser:pgreplparser",
     "//pkg/sql/pgrepl/pgreplparser:pgreplparser_test",
     "//pkg/sql/pgrepl/pgrepltree:pgrepltree",

pkg/acceptance/generated_cli_test.go

@@ -0,0 +1,20 @@
+#! /usr/bin/env expect -f
+
+source [file join [file dirname $argv0] common.tcl]
+
+start_server $argv
+
+start_test "Ensure that replication mode works as expected in the sql shell"
+
+# Spawn a sql shell.
+spawn /bin/bash
+
+send "$argv sql --url `cat server_url`'\&replication=database' -e 'IDENTIFY_SYSTEM'\r"
+eexpect "(1 row)"
+
+send_eof
+eexpect eof
+
+end_test
+
+stop_server $argv

pkg/cli/interactive_tests/test_replication_protocol.tcl

@@ -127,6 +127,7 @@ go_library(
         "grant_revoke_system.go",
         "grant_role.go",
         "group.go",
+        "identify_system.go",
         "index_backfiller.go",
         "index_join.go",
         "index_split_scatter.go",
@@ -442,6 +443,9 @@ go_library(
         "//pkg/sql/paramparse",
         "//pkg/sql/parser",
         "//pkg/sql/parser/statements",
+        "//pkg/sql/pgrepl/lsn",
+        "//pkg/sql/pgrepl/lsnutil",
+        "//pkg/sql/pgrepl/pgrepltree",
         "//pkg/sql/pgwire/pgcode",
         "//pkg/sql/pgwire/pgerror",
         "//pkg/sql/pgwire/pgnotice",

pkg/sql/BUILD.bazel

@@ -344,3 +344,11 @@ const RangesExtraRenders = `
 	coalesce((crdb_internal.range_stats(start_key)->>'range_key_bytes')::INT, 0) +
 	coalesce((crdb_internal.range_stats(start_key)->>'range_val_bytes')::INT, 0) AS range_size
 `
+
+// IdentifySystemColumns is the schema for IDENTIFY_SYSTEM.
+var IdentifySystemColumns = ResultColumns{
+	{Name: "systemid", Typ: types.String},
+	{Name: "timeline", Typ: types.Int4},
+	{Name: "xlogpos", Typ: types.String},
+	{Name: "dbname", Typ: types.String},
+}

pkg/sql/catalog/colinfo/result_columns.go

@@ -53,13 +53,15 @@ go_library(
         "//pkg/sql/oidext",
         "//pkg/sql/opt/cat",
         "//pkg/sql/parser",
+        "//pkg/sql/pgrepl/pgreplparser",
         "//pkg/sql/pgwire/pgcode",
         "//pkg/sql/pgwire/pgerror",
         "//pkg/sql/privilege",
         "//pkg/sql/roleoption",
         "//pkg/sql/sem/catconstants",
         "//pkg/sql/sem/eval",
         "//pkg/sql/sem/tree",
+        "//pkg/sql/sessiondatapb",
         "//pkg/sql/sqlerrors",
         "//pkg/sql/sqltelemetry",
         "//pkg/sql/syntheticprivilege",

pkg/sql/delegate/BUILD.bazel

@@ -18,7 +18,9 @@ import (
 	"github.com/cockroachdb/cockroach/pkg/sql/catalog/colinfo"
 	"github.com/cockroachdb/cockroach/pkg/sql/lexbase"
 	"github.com/cockroachdb/cockroach/pkg/sql/parser"
+	"github.com/cockroachdb/cockroach/pkg/sql/pgrepl/pgreplparser"
 	"github.com/cockroachdb/cockroach/pkg/sql/sem/tree"
+	"github.com/cockroachdb/cockroach/pkg/sql/sessiondatapb"
 )
 
 // delegateShowSyntax implements SHOW SYNTAX. This statement is usually handled
@@ -41,6 +43,16 @@ func (d *delegator) delegateShowSyntax(n *tree.ShowSyntax) (tree.Statement, erro
 		colinfo.ShowSyntaxColumns[0].Name, colinfo.ShowSyntaxColumns[1].Name,
 	)
 
+	// For replication based statements, return nothing for now.
+	if d.evalCtx.SessionData().ReplicationMode != sessiondatapb.ReplicationMode_REPLICATION_MODE_DISABLED &&
+		pgreplparser.IsReplicationProtocolCommand(n.Statement) {
+		return d.parse(fmt.Sprintf(
+			`SELECT '' AS %s, '' AS %s FROM generate_series(0, -1) x`,
+			colinfo.ShowSyntaxColumns[0].Name,
+			colinfo.ShowSyntaxColumns[1].Name,
+		))
+	}
+
 	comma := ""
 	// TODO(knz): in the call below, reportErr is nil although we might
 	// want to be able to capture (and report) these errors as well.

pkg/sql/delegate/show_syntax.go

@@ -2152,6 +2152,7 @@ type SessionArgs struct {
 	User                        username.SQLUsername
 	IsSuperuser                 bool
 	IsSSL                       bool
+	ReplicationMode             sessiondatapb.ReplicationMode
 	SystemIdentity              username.SQLUsername
 	SessionDefaults             SessionDefaults
 	CustomOptionSessionDefaults SessionDefaults

pkg/sql/exec_util.go

@@ -0,0 +1,66 @@
+// Copyright 2023 The Cockroach Authors.
+//
+// Use of this software is governed by the Business Source License
+// included in the file licenses/BSL.txt.
+//
+// As of the Change Date specified in that file, in accordance with
+// the Business Source License, use of this software will be governed
+// by the Apache License, Version 2.0, included in the file
+// licenses/APL.txt.
+
+package sql
+
+import (
+	"context"
+
+	"github.com/cockroachdb/cockroach/pkg/sql/pgrepl/lsn"
+	"github.com/cockroachdb/cockroach/pkg/sql/pgrepl/lsnutil"
+	"github.com/cockroachdb/cockroach/pkg/sql/pgrepl/pgrepltree"
+	"github.com/cockroachdb/cockroach/pkg/sql/sem/tree"
+)
+
+type identifySystemNode struct {
+	optColumnsSlot
+	clusterID string
+	database  string
+	lsn       lsn.LSN
+	shown     bool
+}
+
+func (s *identifySystemNode) startExec(params runParams) error {
+	return nil
+}
+
+func (s *identifySystemNode) Next(params runParams) (bool, error) {
+	if s.shown {
+		return false, nil
+	}
+	s.shown = true
+	return true, nil
+}
+
+func (s *identifySystemNode) Values() tree.Datums {
+	db := tree.DNull
+	if s.database != "" {
+		db = tree.NewDString(s.database)
+	}
+	return tree.Datums{
+		tree.NewDString(s.clusterID),
+		tree.NewDInt(1), // timeline
+		tree.NewDString(s.lsn.String()),
+		db,
+	}
+}
+
+func (s *identifySystemNode) Close(ctx context.Context) {}
+
+func (p *planner) IdentifySystem(
+	ctx context.Context, n *pgrepltree.IdentifySystem,
+) (planNode, error) {
+	return &identifySystemNode{
+		// TODO(#105130): correctly populate this field.
+		lsn:       lsnutil.HLCToLSN(p.Txn().ReadTimestamp()),
+		clusterID: p.ExecCfg().NodeInfo.LogicalClusterID().String(),
+		database:  p.SessionData().Database,
+	}, nil
+}

pkg/sql/identify_system.go

@@ -17,6 +17,7 @@ import (
 	"github.com/cockroachdb/cockroach/pkg/sql/catalog/colinfo"
 	"github.com/cockroachdb/cockroach/pkg/sql/opt"
 	"github.com/cockroachdb/cockroach/pkg/sql/opt/optbuilder"
+	"github.com/cockroachdb/cockroach/pkg/sql/pgrepl/pgrepltree"
 	"github.com/cockroachdb/cockroach/pkg/sql/pgwire/pgcode"
 	"github.com/cockroachdb/cockroach/pkg/sql/pgwire/pgerror"
 	"github.com/cockroachdb/cockroach/pkg/sql/sem/eval"
@@ -281,6 +282,8 @@ func planOpaque(ctx context.Context, p *planner, stmt tree.Statement) (planNode,
 		return p.Truncate(ctx, n)
 	case *tree.Unlisten:
 		return p.Unlisten(ctx, n)
+	case *pgrepltree.IdentifySystem:
+		return p.IdentifySystem(ctx, n)
 	case tree.CCLOnlyStatement:
 		plan, err := p.maybePlanHook(ctx, stmt)
 		if plan == nil && err == nil {
@@ -397,6 +400,8 @@ func init() {
 		&tree.Truncate{},
 		&tree.Unlisten{},
 
+		&pgrepltree.IdentifySystem{},
+
 		// CCL statements (without Export which has an optimizer operator).
 		&tree.AlterBackup{},
 		&tree.AlterBackupSchedule{},