bootstrap: add string representation of initial values

This commit adds InitialValuesToString and InitialValuesFromString which
convert KV datums and KV split keys from a MetadataSchema to a string
representation and vice-versa. The tenant prefix is omitted from the
string representation.

This functionality exists to allow bootstrapping clusters using the
values and splits as generated by an earlier cockroach binary.

Informs cockroachdb#94773.

Release note: None

pkg/BUILD.bazel

@@ -283,6 +283,7 @@ ALL_TESTS = [
     "//pkg/spanconfig:spanconfig_test",
     "//pkg/sql/backfill:backfill_test",
     "//pkg/sql/cacheutil:cacheutil_test",
+    "//pkg/sql/catalog/bootstrap:bootstrap_test",
     "//pkg/sql/catalog/catalogkeys:catalogkeys_test",
     "//pkg/sql/catalog/catenumpb:catenumpb_disallowed_imports_test",
     "//pkg/sql/catalog/catformat:catformat_test",
@@ -1416,6 +1417,7 @@ GO_TARGETS = [
     "//pkg/sql/cacheutil:cacheutil",
     "//pkg/sql/cacheutil:cacheutil_test",
     "//pkg/sql/catalog/bootstrap:bootstrap",
+    "//pkg/sql/catalog/bootstrap:bootstrap_test",
     "//pkg/sql/catalog/catalogkeys:catalogkeys",
     "//pkg/sql/catalog/catalogkeys:catalogkeys_test",
     "//pkg/sql/catalog/catenumpb:catenumpb",

pkg/sql/catalog/bootstrap/BUILD.bazel

@@ -1,5 +1,5 @@
 load("//build/bazelutil/unused_checker:unused.bzl", "get_x_data")
-load("@io_bazel_rules_go//go:def.bzl", "go_library")
+load("@io_bazel_rules_go//go:def.bzl", "go_library", "go_test")
 
 go_library(
     name = "bootstrap",
@@ -31,4 +31,21 @@ go_library(
     ],
 )
 
+go_test(
+    name = "bootstrap_test",
+    srcs = ["bootstrap_test.go"],
+    args = ["-test.timeout=295s"],
+    data = glob(["testdata/**"]),
+    embed = [":bootstrap"],
+    deps = [
+        "//pkg/config/zonepb",
+        "//pkg/keys",
+        "//pkg/roachpb",
+        "//pkg/testutils/datapathutils",
+        "//pkg/util/leaktest",
+        "@com_github_cockroachdb_datadriven//:datadriven",
+        "@com_github_stretchr_testify//require",
+    ],
+)
+
 get_x_data(name = "get_x_data")

pkg/sql/catalog/bootstrap/bootstrap_test.go

@@ -0,0 +1,105 @@
+// Copyright 2022 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 bootstrap
+
+import (
+	"crypto/sha256"
+	"encoding/hex"
+	"testing"
+
+	"github.com/cockroachdb/cockroach/pkg/config/zonepb"
+	"github.com/cockroachdb/cockroach/pkg/keys"
+	"github.com/cockroachdb/cockroach/pkg/roachpb"
+	"github.com/cockroachdb/cockroach/pkg/testutils/datapathutils"
+	"github.com/cockroachdb/cockroach/pkg/util/leaktest"
+	"github.com/cockroachdb/datadriven"
+	"github.com/stretchr/testify/require"
+)
+
+func TestInitialValuesToString(t *testing.T) {
+	defer leaktest.AfterTest(t)()
+	datadriven.Walk(t, datapathutils.TestDataPath(t), func(t *testing.T, path string) {
+		datadriven.RunTest(t, path, func(t *testing.T, d *datadriven.TestData) string {
+			codec := keys.SystemSQLCodec
+			switch d.Cmd {
+			case "system":
+				break
+
+			case "tenant":
+				const dummyTenantID = 12345
+				codec = keys.MakeSQLCodec(roachpb.MustMakeTenantID(dummyTenantID))
+
+			default:
+				t.Fatalf("unexpected command %q", d.Cmd)
+			}
+			var expectedHash string
+			d.ScanArgs(t, "hash", &expectedHash)
+			ms := MakeMetadataSchema(codec, zonepb.DefaultZoneConfigRef(), zonepb.DefaultSystemZoneConfigRef())
+			result := InitialValuesToString(ms)
+			h := sha256.Sum256([]byte(result))
+			if actualHash := hex.EncodeToString(h[:]); expectedHash != actualHash {
+				t.Errorf(`Unexpected hash value %s for %s.
+If you're seeing this error message, this means that the bootstrapped system
+schema has changed. Assuming that this is expected:
+- If this occurred during development on the main branch, rewrite the expected
+  test output and the hash value and move on.
+- If this occurred during development of a patch for a release branch, make
+  very sure that the underlying change really is expected and is backward-
+  compatible and is absolutely necessary. If that's the case, then there are
+  hardcoded literals in the main development branch as well as any subsequent
+  release branches that need to be updated also.`, actualHash, d.Cmd)
+			}
+			return result
+		})
+	})
+}
+
+func TestRoundTripInitialValuesStringRepresentation(t *testing.T) {
+	t.Run("system", func(t *testing.T) {
+		roundTripInitialValuesStringRepresentation(t, 0 /* tenantID */)
+	})
+	t.Run("tenant", func(t *testing.T) {
+		const dummyTenantID = 54321
+		roundTripInitialValuesStringRepresentation(t, dummyTenantID)
+	})
+	t.Run("tenants", func(t *testing.T) {
+		const dummyTenantID1, dummyTenantID2 = 54321, 12345
+		require.Equal(t,
+			InitialValuesToString(makeMetadataSchema(dummyTenantID1)),
+			InitialValuesToString(makeMetadataSchema(dummyTenantID2)),
+		)
+	})
+}
+
+func roundTripInitialValuesStringRepresentation(t *testing.T, tenantID uint64) {
+	ms := makeMetadataSchema(tenantID)
+	expectedKVs, expectedSplits := ms.GetInitialValues()
+	actualKVs, actualSplits, err := InitialValuesFromString(ms.codec, InitialValuesToString(ms))
+	require.NoError(t, err)
+	require.Len(t, actualKVs, len(expectedKVs))
+	require.Len(t, actualSplits, len(expectedSplits))
+	for i, actualKV := range actualKVs {
+		expectedKV := expectedKVs[i]
+		require.EqualValues(t, expectedKV, actualKV)
+	}
+	for i, actualSplit := range actualSplits {
+		expectedSplit := expectedSplits[i]
+		require.EqualValues(t, expectedSplit, actualSplit)
+	}
+}
+
+func makeMetadataSchema(tenantID uint64) MetadataSchema {
+	codec := keys.SystemSQLCodec
+	if tenantID > 0 {
+		codec = keys.MakeSQLCodec(roachpb.MustMakeTenantID(tenantID))
+	}
+	return MakeMetadataSchema(codec, zonepb.DefaultZoneConfigRef(), zonepb.DefaultSystemZoneConfigRef())
+}

pkg/sql/catalog/bootstrap/metadata.go

@@ -13,8 +13,12 @@
 package bootstrap
 
 import (
+	"bytes"
 	"context"
+	"encoding/hex"
+	"encoding/json"
 	"sort"
+	"strings"
 
 	"github.com/cockroachdb/cockroach/pkg/config/zonepb"
 	"github.com/cockroachdb/cockroach/pkg/keys"
@@ -251,6 +255,93 @@ func (ms MetadataSchema) GetInitialValues() ([]roachpb.KeyValue, []roachpb.RKey)
 	return ret, splits
 }
 
+type initialValueStrings struct {
+	Key   string `json:"key"`
+	Value string `json:"value,omitempty"`
+}
+
+// InitialValuesToString returns a string representation of the return values
+// of MetadataSchema.GetInitialValues. The tenant prefix is stripped.
+func InitialValuesToString(ms MetadataSchema) string {
+	kvs, splits := ms.GetInitialValues()
+	// Collect the records.
+	type record struct {
+		k, v []byte
+	}
+	records := make([]record, 0, len(kvs)+len(splits))
+	for _, kv := range kvs {
+		records = append(records, record{k: kv.Key, v: kv.Value.TagAndDataBytes()})
+	}
+	for _, s := range splits {
+		records = append(records, record{k: s})
+	}
+	// Strip the tenant prefix if there is one.
+	p := []byte(ms.codec.TenantPrefix())
+	for i, r := range records {
+		if !bytes.Equal(p, r.k[:len(p)]) {
+			panic("unexpected prefix")
+		}
+		records[i].k = r.k[len(p):]
+	}
+	// Build the string representation.
+	s := make([]initialValueStrings, len(records))
+	for i, r := range records {
+		s[i].Key = hex.EncodeToString(r.k)
+		s[i].Value = hex.EncodeToString(r.v)
+	}
+	// Sort the records by key.
+	sort.Slice(s, func(i, j int) bool {
+		return s[i].Key < s[j].Key
+	})
+	// Build the string representation.
+	var sb strings.Builder
+	sb.WriteRune('[')
+	for i, r := range s {
+		if i > 0 {
+			sb.WriteRune(',')
+		}
+		b, err := json.Marshal(r)
+		if err != nil {
+			panic(err)
+		}
+		sb.Write(b)
+		sb.WriteRune('\n')
+	}
+	sb.WriteRune(']')
+	return sb.String()
+}
+
+// InitialValuesFromString is the reciprocal to InitialValuesToString and
+// appends the tenant prefix from the given codec.
+func InitialValuesFromString(
+	codec keys.SQLCodec, str string,
+) (kvs []roachpb.KeyValue, splits []roachpb.RKey, _ error) {
+	p := codec.TenantPrefix()
+	var s []initialValueStrings
+	if err := json.Unmarshal([]byte(str), &s); err != nil {
+		return nil, nil, err
+	}
+	for i, r := range s {
+		k, err := hex.DecodeString(r.Key)
+		if err != nil {
+			return nil, nil, errors.Errorf("failed to decode hex key %s for record #%d", r.Key, i+1)
+		}
+		v, err := hex.DecodeString(r.Value)
+		if err != nil {
+			return nil, nil, errors.Errorf("failed to decode hex value %s fo record #%d", r.Value, i+1)
+		}
+		k = append(p[:len(p):len(p)], k...)
+		if len(v) == 0 {
+			splits = append(splits, k)
+		} else {
+			kv := roachpb.KeyValue{Key: k}
+			kv.Value.SetTagAndData(v)
+			kvs = append(kvs, kv)
+		}
+	}
+	return kvs, splits, nil
+}
+
 // DescriptorIDs returns the descriptor IDs present in the metadata schema in
 // sorted order.
 func (ms MetadataSchema) DescriptorIDs() descpb.IDs {