Merge cockroachdb#119819 cockroachdb#120384 cockroachdb#120493 cockro…

…achdb#120529

119819: status: export metrics about MemStats into timeseries r=lyang24 a=lyang24

This commit exposes 5 metrics into cockroachdb's RuntimeStatSampler timeseries. The added metrics are MemStackSysBytes, HeapFragmentBytes, HeapReservedBytes, HeapReleasedBytes, TotalAlloc. These metrics are derived from rumtime/metrics.

Fixes: cockroachdb#96717

Relase note: None

120384: sql: add FORCE_INVERTED_INDEX hint r=DrewKimball a=mgartner

#### sql/parser: add parse-no-verify test command

The `parse-no-verify` test command has been added to data-driven parser
tests. In addition, the `error` test command now asserts that parsing
the statement results in an error.

Release note: None

#### sql/parser: move index hint tests to new file

This is purely a mechanical movement of parser tests with index hints
into a new file.

Release note: None

#### sql/parser: support FORCE_INVERTED_INDEX hint

This commit adds parsing support for the `FORCE_INVERTED_INDEX` hint. The
hint currently has no effect.

Release note: None

#### sql: support FORCE_INVERTED_INDEX hint

Epic: None

Release note (sql change): The `FORCE_INVERTED_INDEX` hint is now
supported. This makes the optimizer prefer a query plan scan over any
inverted index of the hinted table. The query will result in an error if
no such query plan can be generated.


120493: drt: cleanup workload runner scripts r=itsbilal a=ajstorm

Cleanup the workload runner scripts to:

- Add the chaos test to systemd-run
- Fix a few errors in cct_tpcc_drop
- Add kv and tpcc scripts

Release note: none
Epic: none

120529: util/mon: reduce sizes of BytesMonitor and BoundAccount structs r=yuzefovich a=yuzefovich

This PR audits BytesMonitor and BoundAccount structs in order to clean up their creation and reduce their memory footprint. In particular, BytesMonitor goes from 208 bytes to 160 bytes and BoundAccount from 32 bytes to 24 bytes.

See each commit for more details.

Microbenchmarks are [here](https://gist.github.com/yuzefovich/f61ad24da47dacfee92dc06dd214d26f) - the improvement is solely due to stack allocating a slice of children in one of the commits.

Inspired by https://github.com/cockroachlabs/support/issues/2847.

Epic: None

Co-authored-by: lyang24 <[email protected]>
Co-authored-by: Marcus Gartner <[email protected]>
Co-authored-by: Adam Storm <[email protected]>
Co-authored-by: Yahor Yuzefovich <[email protected]>

docs/RFCS/20170317_settings_table.md

@@ -37,7 +37,6 @@ might be nice to be able to alter at runtime, without rebooting:
 * `COCKROACH_LEASE_REBALANCING_AGGRESSIVENESS`
 * `COCKROACH_CONSISTENCY_CHECK_INTERVAL`
 * `COCKROACH_MEMORY_ALLOCATION_CHUNK_SIZE`
-* `COCKROACH_NOTEWORTHY_SESSION_MEMORY_USAGE`
 * `COCKROACH_DISABLE_SQL_EVENT_LOG`
 * `COCKROACH_TRACE_SQL`
 

docs/generated/metrics/metrics.html

@@ -1578,6 +1578,11 @@
 <tr><td>SERVER</td><td>sys.gc.pause.ns</td><td>Total GC pause</td><td>GC Pause</td><td>GAUGE</td><td>NANOSECONDS</td><td>AVG</td><td>NONE</td></tr>
 <tr><td>SERVER</td><td>sys.gc.pause.percent</td><td>Current GC pause percentage</td><td>GC Pause</td><td>GAUGE</td><td>PERCENT</td><td>AVG</td><td>NONE</td></tr>
 <tr><td>SERVER</td><td>sys.go.allocbytes</td><td>Current bytes of memory allocated by go</td><td>Memory</td><td>GAUGE</td><td>BYTES</td><td>AVG</td><td>NONE</td></tr>
+<tr><td>SERVER</td><td>sys.go.heap.allocbytes</td><td>Cumulative bytes allocated for heap objects.</td><td>Memory</td><td>GAUGE</td><td>BYTES</td><td>AVG</td><td>NONE</td></tr>
+<tr><td>SERVER</td><td>sys.go.heap.heapfragmentbytes</td><td>Total heap fragmentation bytes, derived from bytes in in-use spans minus bytes allocated</td><td>Memory</td><td>GAUGE</td><td>BYTES</td><td>AVG</td><td>NONE</td></tr>
+<tr><td>SERVER</td><td>sys.go.heap.heapreleasedbytes</td><td>Total bytes returned to the OS from heap.</td><td>Memory</td><td>GAUGE</td><td>BYTES</td><td>AVG</td><td>NONE</td></tr>
+<tr><td>SERVER</td><td>sys.go.heap.heapreservedbytes</td><td>Total bytes reserved by heap, derived from bytes in idle (unused) spans subtracts bytes returned to the OS</td><td>Memory</td><td>GAUGE</td><td>BYTES</td><td>AVG</td><td>NONE</td></tr>
+<tr><td>SERVER</td><td>sys.go.stack.systembytes</td><td>Stack memory obtained from the OS.</td><td>Memory</td><td>GAUGE</td><td>BYTES</td><td>AVG</td><td>NONE</td></tr>
 <tr><td>SERVER</td><td>sys.go.totalbytes</td><td>Total bytes of memory allocated by go, but not released</td><td>Memory</td><td>GAUGE</td><td>BYTES</td><td>AVG</td><td>NONE</td></tr>
 <tr><td>SERVER</td><td>sys.goroutines</td><td>Current number of goroutines</td><td>goroutines</td><td>GAUGE</td><td>COUNT</td><td>AVG</td><td>NONE</td></tr>
 <tr><td>SERVER</td><td>sys.host.disk.io.time</td><td>Time spent reading from or writing to all disks since this process started (as reported by the OS)</td><td>Time</td><td>GAUGE</td><td>NANOSECONDS</td><td>AVG</td><td>NONE</td></tr>

docs/generated/sql/bnf/stmt_block.bnf

@@ -1168,6 +1168,7 @@ unreserved_keyword ::=
 	| 'FORCE_NULL'
 	| 'FORCE_QUOTE'
 	| 'FORCE_INDEX'
+	| 'FORCE_INVERTED_INDEX'
 	| 'FORCE_ZIGZAG'
 	| 'FORWARD'
 	| 'FREEZE'
@@ -3762,6 +3763,7 @@ bare_label_keywords ::=
 	| 'FORCE_NULL'
 	| 'FORCE_QUOTE'
 	| 'FORCE_INDEX'
+	| 'FORCE_INVERTED_INDEX'
 	| 'FORCE_ZIGZAG'
 	| 'FOREIGN'
 	| 'FORMAT'

pkg/base/config.go

@@ -898,15 +898,12 @@ func newTempStorageConfig(
 	} else {
 		monitorName = "temp disk storage"
 	}
-	monitor := mon.NewMonitor(
-		monitorName,
-		mon.DiskResource,
-		nil,             /* curCount */
-		nil,             /* maxHist */
-		1024*1024,       /* increment */
-		maxSizeBytes/10, /* noteworthy */
-		st,
-	)
+	monitor := mon.NewMonitor(mon.Options{
+		Name:      monitorName,
+		Res:       mon.DiskResource,
+		Increment: 1024 * 1024,
+		Settings:  st,
+	})
 	monitor.Start(ctx, nil /* pool */, mon.NewStandaloneBudget(maxSizeBytes))
 	return TempStorageConfig{
 		InMemory: inMemory,

pkg/base/test_server_args.go

@@ -508,15 +508,12 @@ func DefaultTestTempStorageConfig(st *cluster.Settings) TempStorageConfig {
 func DefaultTestTempStorageConfigWithSize(
 	st *cluster.Settings, maxSizeBytes int64,
 ) TempStorageConfig {
-	monitor := mon.NewMonitor(
-		"in-mem temp storage",
-		mon.DiskResource,
-		nil,             /* curCount */
-		nil,             /* maxHist */
-		1024*1024,       /* increment */
-		maxSizeBytes/10, /* noteworthy */
-		st,
-	)
+	monitor := mon.NewMonitor(mon.Options{
+		Name:      "in-mem temp storage",
+		Res:       mon.DiskResource,
+		Increment: 1024 * 1024,
+		Settings:  st,
+	})
 	monitor.Start(context.Background(), nil /* pool */, mon.NewStandaloneBudget(maxSizeBytes))
 	return TempStorageConfig{
 		InMemory: true,

pkg/ccl/backupccl/backup_test.go

@@ -8621,7 +8621,10 @@ func TestReadBackupManifestMemoryMonitoring(t *testing.T) {
 	)
 	require.NoError(t, err)
 
-	m := mon.NewMonitor("test-monitor", mon.MemoryResource, nil, nil, 0, 0, st)
+	m := mon.NewMonitor(mon.Options{
+		Name:     "test-monitor",
+		Settings: st,
+	})
 	m.Start(ctx, nil, mon.NewStandaloneBudget(128<<20))
 	mem := m.MakeBoundAccount()
 	encOpts := &jobspb.BackupEncryptionOptions{

pkg/ccl/backupccl/backuputils/memory_backed_quota_pool_test.go

@@ -22,6 +22,15 @@ import (
 	"github.com/stretchr/testify/require"
 )
 
+func getMemoryMonitor(limit int64) *mon.BytesMonitor {
+	return mon.NewMonitor(mon.Options{
+		Name:      "test-mon",
+		Limit:     limit,
+		Increment: 1,
+		Settings:  cluster.MakeTestingClusterSettings(),
+	})
+}
+
 func TestMemoryBackedQuotaPool(t *testing.T) {
 	defer leaktest.AfterTest(t)()
 	ctx := context.Background()
@@ -75,10 +84,7 @@ func TestMemoryBackedQuotaPool(t *testing.T) {
 	t.Run("external-mon-user", func(t *testing.T) {
 		// Create a quota pool of limit 10.
 		const limit = 10
-		mm := mon.NewMonitorWithLimit(
-			"test-mon", mon.MemoryResource, limit,
-			nil, nil, 1, 0,
-			cluster.MakeTestingClusterSettings())
+		mm := getMemoryMonitor(limit)
 		mm.Start(ctx, nil, mon.NewStandaloneBudget(limit))
 		defer mm.Stop(ctx)
 
@@ -129,10 +135,7 @@ func TestMemoryBackedQuotaPoolConcurrent(t *testing.T) {
 		for _, numGoroutines := range []int{1, 10, 100} {
 			quota := quota
 			ctx := context.Background()
-			mm := mon.NewMonitorWithLimit(
-				"test-mon", mon.MemoryResource, quota,
-				nil, nil, 1, 0,
-				cluster.MakeTestingClusterSettings())
+			mm := getMemoryMonitor(quota)
 			mm.Start(ctx, nil, mon.NewStandaloneBudget(quota))
 			mem := mm.MakeConcurrentBoundAccount()
 
@@ -187,10 +190,7 @@ func TestMemoryBackedQuotaPoolConcurrent(t *testing.T) {
 func makeTestingQuotaPool(
 	ctx context.Context, limit int64,
 ) (qp *MemoryBackedQuotaPool, cleanup func()) {
-	mm := mon.NewMonitorWithLimit(
-		"test-mon", mon.MemoryResource, limit,
-		nil, nil, 1, 0,
-		cluster.MakeTestingClusterSettings())
+	mm := getMemoryMonitor(limit)
 	mm.Start(ctx, nil, mon.NewStandaloneBudget(limit))
 
 	qp = NewMemoryBackedQuotaPool(ctx, mm, "test-qp", limit)

pkg/ccl/backupccl/restore_data_processor_test.go

@@ -271,9 +271,12 @@ func runTestIngest(t *testing.T, init func(*cluster.Settings)) {
 					cloud.NilMetrics,
 					opts...)
 			},
-			Settings:          s.ClusterSettings(),
-			Codec:             s.Codec(),
-			BackupMonitor:     mon.NewUnlimitedMonitor(ctx, "test", mon.MemoryResource, nil, nil, 0, s.ClusterSettings()),
+			Settings: s.ClusterSettings(),
+			Codec:    s.Codec(),
+			BackupMonitor: mon.NewUnlimitedMonitor(ctx, mon.Options{
+				Name:     "test",
+				Settings: s.ClusterSettings(),
+			}),
 			BulkSenderLimiter: limit.MakeConcurrentRequestLimiter("test", math.MaxInt),
 		},
 		EvalCtx: &eval.Context{

pkg/ccl/changefeedccl/changefeed_test.go

@@ -7938,11 +7938,12 @@ func TestChangefeedPredicateWithSchemaChange(t *testing.T) {
 }
 
 func startMonitorWithBudget(budget int64) *mon.BytesMonitor {
-	mm := mon.NewMonitorWithLimit(
-		"test-mm", mon.MemoryResource, budget,
-		nil, nil,
-		128 /* small allocation increment */, 100,
-		cluster.MakeTestingClusterSettings())
+	mm := mon.NewMonitor(mon.Options{
+		Name:      "test-mm",
+		Limit:     budget,
+		Increment: 128, /* small allocation increment */
+		Settings:  cluster.MakeTestingClusterSettings(),
+	})
 	mm.Start(context.Background(), nil, mon.NewStandaloneBudget(budget))
 	return mm
 }

pkg/ccl/changefeedccl/kvevent/blocking_buffer_test.go

@@ -70,11 +70,12 @@ func makeRangeFeedEvent(rnd *rand.Rand, valSize int, prevValSize int) *kvpb.Rang
 }
 
 func getBoundAccountWithBudget(budget int64) (account mon.BoundAccount, cleanup func()) {
-	mm := mon.NewMonitorWithLimit(
-		"test-mm", mon.MemoryResource, budget,
-		nil, nil,
-		128 /* small allocation increment */, 100,
-		cluster.MakeTestingClusterSettings())
+	mm := mon.NewMonitor(mon.Options{
+		Name:      "test-mm",
+		Limit:     budget,
+		Increment: 128, /* small allocation increment */
+		Settings:  cluster.MakeTestingClusterSettings(),
+	})
 	mm.Start(context.Background(), nil, mon.NewStandaloneBudget(budget))
 	return mm.MakeBoundAccount(), func() { mm.Stop(context.Background()) }
 }

pkg/ccl/changefeedccl/kvfeed/kv_feed_test.go

@@ -117,10 +117,10 @@ func TestKVFeed(t *testing.T) {
 	st := cluster.MakeTestingClusterSettings()
 	runTest := func(t *testing.T, tc testCase) {
 		settings := cluster.MakeTestingClusterSettings()
-		mm := mon.NewUnlimitedMonitor(
-			context.Background(), "test", mon.MemoryResource,
-			nil /* curCount */, nil /* maxHist */, math.MaxInt64, settings,
-		)
+		mm := mon.NewUnlimitedMonitor(context.Background(), mon.Options{
+			Name:     "test",
+			Settings: settings,
+		})
 		metrics := kvevent.MakeMetrics(time.Minute)
 		buf := kvevent.NewMemBuffer(mm.MakeBoundAccount(), &st.SV, &metrics)
 

pkg/cmd/drt/scripts/cct_tpcc_drop.sh

@@ -1,8 +1,14 @@
 #!/usr/bin/env bash
 
 CLUSTER_NAME=cct-232
-PG_URL_N1=$(roachprod pgurl $CLUSTER_NAME:1 --secure)
-PG_URLS=$(roachprod pgurl $CLUSTER_NAME --secure)
+PG_URL_N1=$(./roachprod pgurl $CLUSTER_NAME:1 --secure --cluster=application --external | sed s/\'//g)
+PGURLS=$(./roachprod pgurl cct-232 --external --secure --cluster application | sed s/\'//g)
+
+read -r -a PGURLS_ARR <<< "$PGURLS"
+
+echo ">> Dropping old databases if they exist"
+./cockroach sql --url "${PG_URL_N1}" -e "DROP DATABASE IF EXISTS cct_tpcc_drop_old CASCADE;"
+./cockroach sql --url "${PG_URL_N1}" -e "DROP DATABASE IF EXISTS cct_tpcc_drop CASCADE;"
 
 j=0
 while true; do
@@ -18,7 +24,7 @@ while true; do
         --db cct_tpcc_drop \
         $PG_URL_N1 | tee "$INIT_LOG"
     echo ">> Dropping cct_tpcc_drop_old if it exists"
-    ./cockroach sql --url $PG_URL_N1 -e "DROP DATABASE cct_tpcc_drop_old CASCADE;"
+    ./cockroach sql --url "${PG_URL_N1}" -e "DROP DATABASE cct_tpcc_drop_old CASCADE;"
     sleep 5
     echo ">> Starting tpcc workload for 1h"
     ./cockroach workload run tpcc \
@@ -31,10 +37,10 @@ while true; do
         --display-every 5s \
         --duration 60m \
   --tolerate-errors \
-   $PG_URLS| tee "$RUN_LOG"
+   "${PGURLS_ARR[@]}" | tee "$RUN_LOG"
 
     echo ">> Renaming to cct_tpcc_drop_old"
-    ./cockroach sql --url $PG_URLS -e "ALTER DATABASE cct_tpcc_drop RENAME TO cct_tpcc_drop_old;"
+    ./cockroach sql --url "${PG_URL_N1}" -e "ALTER DATABASE cct_tpcc_drop RENAME TO cct_tpcc_drop_old;"
 
     sleep 1
 done

pkg/cmd/drt/scripts/kv_run.sh

@@ -0,0 +1,39 @@
+#!/usr/bin/env bash
+
+set -o pipefail
+
+PGURLS=$(./roachprod pgurl cct-232 --external --secure --cluster application | sed s/\'//g)
+
+read -r -a PGURLS_ARR <<< "$PGURLS"
+
+j=0
+while true; do
+    echo ">> Starting kv workload"
+    ((j++))
+    LOG=./kv_$j.txt
+    ./cockroach workload run kv \
+        --init \
+        --drop \
+        --concurrency 128 \
+        --histograms kv/stats.json \
+        --db kv \
+        --splits 1000 \
+        --read-percent 50 \
+        --span-percent 20 \
+        --cycle-length 100000 \
+        --min-block-bytes 100 \
+        --max-block-bytes 1000 \
+        --prometheus-port 2114 \
+        --max-rate 1200 \
+        --secure \
+        --ramp 10m \
+        --display-every 5s \
+        --duration 12h \
+        --tolerate-errors \
+        --enum \
+        $PGURLS | tee "$LOG"
+    if [ $? -eq 0 ]; then
+        rm "$LOG"
+    fi
+    sleep 1
+done

pkg/cmd/drt/scripts/schemachange_run.sh

@@ -0,0 +1,26 @@
+#!/usr/bin/env bash
+set -o pipefail
+
+PGURLS=$(./roachprod pgurl cct-232 --external --secure --cluster application | sed s/\'//g)
+read -r -a PGURLS_ARR <<< "$PGURLS"
+
+j=0
+while true; do
+    echo ">> Starting random schema change workload"
+    ((j++))
+    LOG=./schemachange_$j.txt
+    ./workload run schemachange --verbose=1 \
+         --tolerate-errors=true \
+         --histograms /mnt/data1/schemachange_perf/stats.json \
+         --concurrency 10 \
+         --txn-log /mnt/data1/schemachange_perf/txn.log \
+         --secure \
+         --user cct_schemachange_user \
+         --db schemachange \
+         --password cct_schemachange_password \
+          "${PGURLS_ARR[@]}" | tee "$LOG"
+    if [ 0 -eq 0 ]; then
+        rm "$LOG"
+    fi
+    sleep 1
+done

pkg/cmd/drt/scripts/start_all_workloads.sh

@@ -1,4 +1,6 @@
-sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_kv ./kv_run.sh
-sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_tpcc ./tpcc_run.sh
-sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_tpcc_drop ./cct_tpcc_drop.sh
-sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_schemachange ./schemachange_run.sh
+sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_kv --uid 1000 --gid 1000 ./kv_run.sh
+sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_tpcc --uid 1000 --gid 1000 ./tpcc_run.sh
+sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_tpcc_drop --uid 1000 --gid 1000 ./cct_tpcc_drop.sh
+sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit cct_schemachange --uid 1000 --gid 1000 ./schemachange_run.sh
+sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit chaos_test --uid=1000 --gid=1000 ./chaos_helper.sh
+sudo systemd-run --working-directory=/home/ubuntu --service-type exec --collect --unit roachtest_operations --uid 1000 --gid 1000 ./roachtest_operations_run.sh

pkg/cmd/drt/scripts/stop_all_workloads.sh

@@ -2,3 +2,4 @@ sudo systemctl stop cct_schemachange
 sudo systemctl stop cct_kv
 sudo systemctl stop cct_tpcc
 sudo systemctl stop cct_tpcc_drop
+sudo systemctl stop chaos_test

pkg/cmd/drt/scripts/tpcc_run.sh

@@ -0,0 +1,36 @@
+#!/usr/bin/env bash
+
+set -o pipefail
+
+TPCC_DB=cct_tpcc
+TPCC_USER=cct_tpcc_user
+TPCC_PASSWORD=tpcc
+PGURLS=$(./roachprod pgurl cct-232 --external --secure --cluster application | sed s/\'//g)
+
+read -r -a PGURLS_ARR <<< "$PGURLS"
+
+j=0
+while true; do
+    echo ">> Starting tpcc workload"
+    ((j++))
+    LOG=./tpcc_$j.txt
+    ./cockroach workload run tpcc \
+      --warehouses 3000 \
+      --active-warehouses 1500 \
+      --concurrency 128 \
+      --max-rate 7000 \
+      --db cct_tpcc \
+      --secure \
+      --ramp 10m \
+      --display-every 5s \
+      --duration 12h \
+      --user cct_tpcc_user \
+      --tolerate-errors \
+      --password tpcc \
+      --families \
+        "${PGURLS_ARR[@]}" | tee $LOG
+    if [ $? -eq 0 ]; then
+        rm "$LOG"
+    fi
+    sleep 1
+done

pkg/kv/bulk/buffering_adder.go

@@ -70,7 +70,7 @@ type BufferingAdder struct {
 	importEpoch uint32
 
 	bulkMon *mon.BytesMonitor
-	memAcc  mon.BoundAccount
+	memAcc  mon.EarmarkedBoundAccount
 
 	onFlush func(summary kvpb.BulkOpSummary)
 	// underfill tracks how much capacity was remaining in curBuf when it was
@@ -139,7 +139,7 @@ func MakeBulkAdder(
 	//
 	// TODO(adityamaru): IMPORT should also reserve memory for a single SST which
 	// it will store in-memory before sending it to RocksDB.
-	b.memAcc = bulkMon.MakeBoundAccount()
+	b.memAcc = bulkMon.MakeEarmarkedBoundAccount()
 	if opts.MinBufferSize > 0 {
 		if err := b.memAcc.Reserve(ctx, opts.MinBufferSize); err != nil {
 			return nil, errors.WithHint(

pkg/kv/bulk/kv_buf.go

@@ -50,7 +50,9 @@ const (
 	maxLen, maxOffset = lenMask, 1<<(64-lenBits) - 1
 )
 
-func (b *kvBuf) fits(ctx context.Context, toAdd sz, maxUsed sz, acc *mon.BoundAccount) bool {
+func (b *kvBuf) fits(
+	ctx context.Context, toAdd sz, maxUsed sz, acc *mon.EarmarkedBoundAccount,
+) bool {
 	if len(b.entries) < cap(b.entries) && sz(len(b.slab))+toAdd < sz(cap(b.slab)) {
 		return true // fits in current cap, nothing to do.
 	}