diff --git a/lib/logstorage/parser.go b/lib/logstorage/parser.go
index 55825dae7..210aae97b 100644
--- a/lib/logstorage/parser.go
+++ b/lib/logstorage/parser.go
@@ -247,7 +247,7 @@ func (q *Query) AddTimeFilter(start, end int64) {
 // See https://docs.victoriametrics.com/victorialogs/logsql/#limit-pipe
 func (q *Query) AddPipeLimit(n uint64) {
 	q.pipes = append(q.pipes, &pipeLimit{
-		n: n,
+		limit: n,
 	})
 }
 
@@ -255,6 +255,7 @@ func (q *Query) AddPipeLimit(n uint64) {
 func (q *Query) Optimize() {
 	q.pipes = optimizeSortOffsetPipes(q.pipes)
 	q.pipes = optimizeSortLimitPipes(q.pipes)
+	q.pipes = optimizeUniqLimitPipes(q.pipes)
 	q.pipes = optimizeFilterPipes(q.pipes)
 
 	// Merge `q | filter ...` into q.
@@ -266,6 +267,14 @@ func (q *Query) Optimize() {
 		}
 	}
 
+	// Optimize `q | field_names ...` by marking pipeFieldNames as first pipe
+	if len(q.pipes) > 0 {
+		pf, ok := q.pipes[0].(*pipeFieldNames)
+		if ok {
+			pf.isFirstPipe = true
+		}
+	}
+
 	// Optimize 'in(query)' filters
 	optimizeFilterIn(q.f)
 	for _, p := range q.pipes {
@@ -306,7 +315,7 @@ func optimizeSortOffsetPipes(pipes []pipe) []pipe {
 			continue
 		}
 		if ps.offset == 0 && ps.limit == 0 {
-			ps.offset = po.n
+			ps.offset = po.offset
 		}
 		pipes = append(pipes[:i], pipes[i+1:]...)
 	}
@@ -327,8 +336,30 @@ func optimizeSortLimitPipes(pipes []pipe) []pipe {
 			i++
 			continue
 		}
-		if ps.limit == 0 || pl.n < ps.limit {
-			ps.limit = pl.n
+		if ps.limit == 0 || pl.limit < ps.limit {
+			ps.limit = pl.limit
+		}
+		pipes = append(pipes[:i], pipes[i+1:]...)
+	}
+	return pipes
+}
+
+func optimizeUniqLimitPipes(pipes []pipe) []pipe {
+	// Merge 'uniq ... | limit ...' into 'uniq ... limit ...'
+	i := 1
+	for i < len(pipes) {
+		pl, ok := pipes[i].(*pipeLimit)
+		if !ok {
+			i++
+			continue
+		}
+		pu, ok := pipes[i-1].(*pipeUniq)
+		if !ok {
+			i++
+			continue
+		}
+		if pu.limit == 0 || pl.limit < pu.limit {
+			pu.limit = pl.limit
 		}
 		pipes = append(pipes[:i], pipes[i+1:]...)
 	}
diff --git a/lib/logstorage/parser_test.go b/lib/logstorage/parser_test.go
index 17f93c843..4d7fbeaf1 100644
--- a/lib/logstorage/parser_test.go
+++ b/lib/logstorage/parser_test.go
@@ -811,6 +811,10 @@ func TestParseQuerySuccess(t *testing.T) {
 	// multiple fields pipes
 	f(`foo | fields bar | fields baz, abc`, `foo | fields bar | fields baz, abc`)
 
+	// field_names pipe
+	f(`foo | field_names as x`, `foo | field_names as x`)
+	f(`foo | field_names y`, `foo | field_names as y`)
+
 	// copy and cp pipe
 	f(`* | copy foo as bar`, `* | copy foo as bar`)
 	f(`* | cp foo bar`, `* | copy foo as bar`)
@@ -1210,6 +1214,13 @@ func TestParseQueryFailure(t *testing.T) {
 	f(`foo | fields bar,`)
 	f(`foo | fields bar,,`)
 
+	// invalid field_names
+	f(`foo | field_names ()`)
+	f(`foo | field_names (x)`)
+	f(`foo | field_names (x,y)`)
+	f(`foo | field_names x y`)
+	f(`foo | field_names x, y`)
+
 	// invalid copy and cp pipe
 	f(`foo | copy`)
 	f(`foo | cp`)
@@ -1375,8 +1386,9 @@ func TestQueryGetNeededColumns(t *testing.T) {
 
 		q, err := ParseQuery(s)
 		if err != nil {
-			t.Fatalf("cannot parse query %s: %s", s, err)
+			t.Fatalf("cannot parse query [%s]: %s", s, err)
 		}
+		q.Optimize()
 
 		needed, unneeded := q.getNeededColumns()
 		neededColumns := strings.Join(needed, ",")
@@ -1486,6 +1498,27 @@ func TestQueryGetNeededColumns(t *testing.T) {
 	f(`* | uniq by (f1,f2) | rm f1,f3`, `f1,f2`, ``)
 	f(`* | uniq by (f1,f2) | fields f3`, `f1,f2`, ``)
 
+	f(`* | filter foo f1:bar`, `*`, ``)
+	f(`* | filter foo f1:bar | fields f2`, `f2`, ``)
+	f(`* | limit 10 | filter foo f1:bar | fields f2`, `_msg,f1,f2`, ``)
+	f(`* | filter foo f1:bar | fields f1`, `f1`, ``)
+	f(`* | filter foo f1:bar | rm f1`, `*`, `f1`)
+	f(`* | limit 10 | filter foo f1:bar | rm f1`, `*`, ``)
+	f(`* | filter foo f1:bar | rm f2`, `*`, `f2`)
+	f(`* | limit 10 | filter foo f1:bar | rm f2`, `*`, `f2`)
+	f(`* | fields x | filter foo f1:bar | rm f2`, `x`, ``)
+	f(`* | fields x,f1 | filter foo f1:bar | rm f2`, `f1,x`, ``)
+	f(`* | rm x,f1 | filter foo f1:bar`, `*`, `f1,x`)
+
+	f(`* | field_names as foo`, `*`, `_time`)
+	f(`* | field_names foo | fields bar`, `*`, `_time`)
+	f(`* | field_names foo | fields foo`, `*`, `_time`)
+	f(`* | field_names foo | rm foo`, `*`, `_time`)
+	f(`* | field_names foo | rm bar`, `*`, `_time`)
+	f(`* | field_names foo | rm _time`, `*`, `_time`)
+	f(`* | fields x,y | field_names as bar | fields baz`, `x,y`, ``)
+	f(`* | rm x,y | field_names as bar | fields baz`, `*`, `x,y`)
+
 	f(`* | rm f1, f2`, `*`, `f1,f2`)
 	f(`* | rm f1, f2 | mv f2 f3`, `*`, `f1,f2,f3`)
 	f(`* | rm f1, f2 | cp f2 f3`, `*`, `f1,f2,f3`)
diff --git a/lib/logstorage/pipe.go b/lib/logstorage/pipe.go
index 200f48d36..36adb7b1e 100644
--- a/lib/logstorage/pipe.go
+++ b/lib/logstorage/pipe.go
@@ -83,6 +83,12 @@ func parsePipes(lex *lexer) ([]pipe, error) {
 				return nil, fmt.Errorf("cannot parse 'delete' pipe: %w", err)
 			}
 			pipes = append(pipes, pd)
+		case lex.isKeyword("field_names"):
+			pf, err := parsePipeFieldNames(lex)
+			if err != nil {
+				return nil, fmt.Errorf("cannot parse 'field_names' pipe: %w", err)
+			}
+			pipes = append(pipes, pf)
 		case lex.isKeyword("fields"):
 			pf, err := parsePipeFields(lex)
 			if err != nil {
diff --git a/lib/logstorage/pipe_field_names.go b/lib/logstorage/pipe_field_names.go
new file mode 100644
index 000000000..38a1e265a
--- /dev/null
+++ b/lib/logstorage/pipe_field_names.go
@@ -0,0 +1,161 @@
+package logstorage
+
+import (
+	"fmt"
+	"strings"
+	"unsafe"
+)
+
+// pipeFieldNames processes '| field_names' pipe.
+//
+// See https://docs.victoriametrics.com/victorialogs/logsql/#field-names-pipe
+type pipeFieldNames struct {
+	// resultName is the name of the column to write results to.
+	resultName string
+
+	// isFirstPipe is set to true if '| field_names' pipe is the first in the query.
+	//
+	// This allows skipping loading of _time column.
+	isFirstPipe bool
+}
+
+func (pf *pipeFieldNames) String() string {
+	return "field_names as " + quoteTokenIfNeeded(pf.resultName)
+}
+
+func (pf *pipeFieldNames) updateNeededFields(neededFields, unneededFields fieldsSet) {
+	neededFields.add("*")
+	unneededFields.reset()
+
+	if pf.isFirstPipe {
+		unneededFields.add("_time")
+	}
+}
+
+func (pf *pipeFieldNames) newPipeProcessor(workersCount int, stopCh <-chan struct{}, _ func(), ppBase pipeProcessor) pipeProcessor {
+	shards := make([]pipeFieldNamesProcessorShard, workersCount)
+	for i := range shards {
+		shards[i] = pipeFieldNamesProcessorShard{
+			pipeFieldNamesProcessorShardNopad: pipeFieldNamesProcessorShardNopad{
+				m: make(map[string]struct{}),
+			},
+		}
+	}
+
+	pfp := &pipeFieldNamesProcessor{
+		pf:     pf,
+		stopCh: stopCh,
+		ppBase: ppBase,
+
+		shards: shards,
+	}
+	return pfp
+}
+
+type pipeFieldNamesProcessor struct {
+	pf     *pipeFieldNames
+	stopCh <-chan struct{}
+	ppBase pipeProcessor
+
+	shards []pipeFieldNamesProcessorShard
+}
+
+type pipeFieldNamesProcessorShard struct {
+	pipeFieldNamesProcessorShardNopad
+
+	// The padding prevents false sharing on widespread platforms with 128 mod (cache line size) = 0 .
+	_ [128 - unsafe.Sizeof(pipeFieldNamesProcessorShardNopad{})%128]byte
+}
+
+type pipeFieldNamesProcessorShardNopad struct {
+	// m holds unique field names.
+	m map[string]struct{}
+}
+
+func (pfp *pipeFieldNamesProcessor) writeBlock(workerID uint, br *blockResult) {
+	if len(br.timestamps) == 0 {
+		return
+	}
+
+	shard := &pfp.shards[workerID]
+	cs := br.getColumns()
+	for _, c := range cs {
+		if _, ok := shard.m[c.name]; !ok {
+			nameCopy := strings.Clone(c.name)
+			shard.m[nameCopy] = struct{}{}
+		}
+	}
+}
+
+func (pfp *pipeFieldNamesProcessor) flush() error {
+	if needStop(pfp.stopCh) {
+		return nil
+	}
+
+	// merge state across shards
+	shards := pfp.shards
+	m := shards[0].m
+	shards = shards[1:]
+	for i := range shards {
+		for k := range shards[i].m {
+			m[k] = struct{}{}
+		}
+	}
+
+	// write result
+	wctx := &pipeFieldNamesWriteContext{
+		pfp: pfp,
+	}
+	wctx.rcs[0].name = pfp.pf.resultName
+	for k := range m {
+		wctx.writeRow(k)
+	}
+	wctx.flush()
+
+	return nil
+}
+
+type pipeFieldNamesWriteContext struct {
+	pfp *pipeFieldNamesProcessor
+	rcs [1]resultColumn
+	br  blockResult
+
+	valuesLen int
+}
+
+func (wctx *pipeFieldNamesWriteContext) writeRow(v string) {
+	wctx.rcs[0].addValue(v)
+	wctx.valuesLen += len(v)
+	if wctx.valuesLen >= 1_000_000 {
+		wctx.flush()
+	}
+}
+
+func (wctx *pipeFieldNamesWriteContext) flush() {
+	br := &wctx.br
+
+	wctx.valuesLen = 0
+
+	// Flush rcs to ppBase
+	br.setResultColumns(wctx.rcs[:1])
+	wctx.pfp.ppBase.writeBlock(0, br)
+	br.reset()
+	wctx.rcs[0].resetKeepName()
+}
+
+func parsePipeFieldNames(lex *lexer) (*pipeFieldNames, error) {
+	if !lex.isKeyword("field_names") {
+		return nil, fmt.Errorf("expecting 'field_names'; got %q", lex.token)
+	}
+	lex.nextToken()
+
+	if lex.isKeyword("as") {
+		lex.nextToken()
+	}
+	resultName := getCompoundPhrase(lex, false)
+
+	pf := &pipeFieldNames{
+		resultName: resultName,
+	}
+	return pf, nil
+}
diff --git a/lib/logstorage/pipe_field_names_test.go b/lib/logstorage/pipe_field_names_test.go
new file mode 100644
index 000000000..df8d68c48
--- /dev/null
+++ b/lib/logstorage/pipe_field_names_test.go
@@ -0,0 +1,38 @@
+package logstorage
+
+import (
+	"testing"
+)
+
+func TestPipeFieldNamesUpdateNeededFields(t *testing.T) {
+	f := func(s string, neededFields, unneededFields, neededFieldsExpected, unneededFieldsExpected string) {
+		t.Helper()
+
+		nfs := newTestFieldsSet(neededFields)
+		unfs := newTestFieldsSet(unneededFields)
+
+		lex := newLexer(s)
+		p, err := parsePipeFieldNames(lex)
+		if err != nil {
+			t.Fatalf("cannot parse %s: %s", s, err)
+		}
+		p.updateNeededFields(nfs, unfs)
+
+		assertNeededFields(t, nfs, unfs, neededFieldsExpected, unneededFieldsExpected)
+	}
+
+	// all the needed fields
+	f("field_names as f1", "*", "", "*", "")
+
+	// all the needed fields, unneeded fields do not intersect with src
+	f("field_names as f3", "*", "f1,f2", "*", "")
+
+	// all the needed fields, unneeded fields intersect with src
+	f("field_names as f1", "*", "s1,f1,f2", "*", "")
+
+	// needed fields do not intersect with src
+	f("field_names as f3", "f1,f2", "", "*", "")
+
+	// needed fields intersect with src
+	f("field_names as f1", "s1,f1,f2", "", "*", "")
+}
diff --git a/lib/logstorage/pipe_filter.go b/lib/logstorage/pipe_filter.go
index a3070754d..a8f8787b6 100644
--- a/lib/logstorage/pipe_filter.go
+++ b/lib/logstorage/pipe_filter.go
@@ -17,19 +17,20 @@ func (pf *pipeFilter) String() string {
 	return "filter " + pf.f.String()
 }
 
-func (pf *pipeFilter) updateNeededFields(neededFields, _ fieldsSet) {
-	pf.f.updateNeededFields(neededFields)
+func (pf *pipeFilter) updateNeededFields(neededFields, unneededFields fieldsSet) {
+	if neededFields.contains("*") {
+		fs := newFieldsSet()
+		pf.f.updateNeededFields(fs)
+		for f := range fs {
+			unneededFields.remove(f)
+		}
+	} else {
+		pf.f.updateNeededFields(neededFields)
+	}
 }
 
 func (pf *pipeFilter) newPipeProcessor(workersCount int, _ <-chan struct{}, _ func(), ppBase pipeProcessor) pipeProcessor {
 	shards := make([]pipeFilterProcessorShard, workersCount)
-	for i := range shards {
-		shards[i] = pipeFilterProcessorShard{
-			pipeFilterProcessorShardNopad: pipeFilterProcessorShardNopad{
-				pf: pf,
-			},
-		}
-	}
 
 	pfp := &pipeFilterProcessor{
 		pf:     pf,
@@ -55,8 +56,6 @@ type pipeFilterProcessorShard struct {
 }
 
 type pipeFilterProcessorShardNopad struct {
-	pf *pipeFilter
-
 	br blockResult
 	bm bitmap
 }
@@ -71,7 +70,7 @@ func (pfp *pipeFilterProcessor) writeBlock(workerID uint, br *blockResult) {
 	bm := &shard.bm
 	bm.init(len(br.timestamps))
 	bm.setBits()
-	shard.pf.f.applyToBlockResult(br, bm)
+	pfp.pf.f.applyToBlockResult(br, bm)
 	if bm.areAllBitsSet() {
 		// Fast path - the filter didn't filter out anything - send br to the base pipe as is.
 		pfp.ppBase.writeBlock(workerID, br)
diff --git a/lib/logstorage/pipe_filter_test.go b/lib/logstorage/pipe_filter_test.go
new file mode 100644
index 000000000..d4d5e3ff9
--- /dev/null
+++ b/lib/logstorage/pipe_filter_test.go
@@ -0,0 +1,38 @@
+package logstorage
+
+import (
+	"testing"
+)
+
+func TestPipeFilterUpdateNeededFields(t *testing.T) {
+	f := func(s string, neededFields, unneededFields, neededFieldsExpected, unneededFieldsExpected string) {
+		t.Helper()
+
+		nfs := newTestFieldsSet(neededFields)
+		unfs := newTestFieldsSet(unneededFields)
+
+		lex := newLexer(s)
+		p, err := parsePipeFilter(lex)
+		if err != nil {
+			t.Fatalf("cannot parse %s: %s", s, err)
+		}
+		p.updateNeededFields(nfs, unfs)
+
+		assertNeededFields(t, nfs, unfs, neededFieldsExpected, unneededFieldsExpected)
+	}
+
+	// all the needed fields
+	f("filter foo f1:bar", "*", "", "*", "")
+
+	// all the needed fields, unneeded fields do not intersect with src
+	f("filter foo f3:bar", "*", "f1,f2", "*", "f1,f2")
+
+	// all the needed fields, unneeded fields intersect with src
+	f("filter foo f1:bar", "*", "s1,f1,f2", "*", "s1,f2")
+
+	// needed fields do not intersect with src
+	f("filter foo f3:bar", "f1,f2", "", "_msg,f1,f2,f3", "")
+
+	// needed fields intersect with src
+	f("filter foo f1:bar", "s1,f1,f2", "", "_msg,f1,f2,s1", "")
+}
diff --git a/lib/logstorage/pipe_limit.go b/lib/logstorage/pipe_limit.go
index 91793b9fb..15e0f4f26 100644
--- a/lib/logstorage/pipe_limit.go
+++ b/lib/logstorage/pipe_limit.go
@@ -9,18 +9,18 @@ import (
 //
 // See https://docs.victoriametrics.com/victorialogs/logsql/#limit-pipe
 type pipeLimit struct {
-	n uint64
+	limit uint64
 }
 
 func (pl *pipeLimit) String() string {
-	return fmt.Sprintf("limit %d", pl.n)
+	return fmt.Sprintf("limit %d", pl.limit)
 }
 
 func (pl *pipeLimit) updateNeededFields(_, _ fieldsSet) {
 }
 
 func (pl *pipeLimit) newPipeProcessor(_ int, _ <-chan struct{}, cancel func(), ppBase pipeProcessor) pipeProcessor {
-	if pl.n == 0 {
+	if pl.limit == 0 {
 		// Special case - notify the caller to stop writing data to the returned pipeLimitProcessor
 		cancel()
 	}
@@ -45,7 +45,7 @@ func (plp *pipeLimitProcessor) writeBlock(workerID uint, br *blockResult) {
 	}
 
 	rowsProcessed := plp.rowsProcessed.Add(uint64(len(br.timestamps)))
-	if rowsProcessed <= plp.pl.n {
+	if rowsProcessed <= plp.pl.limit {
 		// Fast path - write all the rows to ppBase.
 		plp.ppBase.writeBlock(workerID, br)
 		return
@@ -53,13 +53,13 @@ func (plp *pipeLimitProcessor) writeBlock(workerID uint, br *blockResult) {
 
 	// Slow path - overflow. Write the remaining rows if needed.
 	rowsProcessed -= uint64(len(br.timestamps))
-	if rowsProcessed >= plp.pl.n {
+	if rowsProcessed >= plp.pl.limit {
 		// Nothing to write. There is no need in cancel() call, since it has been called by another goroutine.
 		return
 	}
 
 	// Write remaining rows.
-	keepRows := plp.pl.n - rowsProcessed
+	keepRows := plp.pl.limit - rowsProcessed
 	br.truncateRows(int(keepRows))
 	plp.ppBase.writeBlock(workerID, br)
 
@@ -83,7 +83,7 @@ func parsePipeLimit(lex *lexer) (*pipeLimit, error) {
 	}
 	lex.nextToken()
 	pl := &pipeLimit{
-		n: n,
+		limit: n,
 	}
 	return pl, nil
 }
diff --git a/lib/logstorage/pipe_offset.go b/lib/logstorage/pipe_offset.go
index 61219989a..5518d80f8 100644
--- a/lib/logstorage/pipe_offset.go
+++ b/lib/logstorage/pipe_offset.go
@@ -9,11 +9,11 @@ import (
 //
 // See https://docs.victoriametrics.com/victorialogs/logsql/#offset-pipe
 type pipeOffset struct {
-	n uint64
+	offset uint64
 }
 
 func (po *pipeOffset) String() string {
-	return fmt.Sprintf("offset %d", po.n)
+	return fmt.Sprintf("offset %d", po.offset)
 }
 
 func (po *pipeOffset) updateNeededFields(_, _ fieldsSet) {
@@ -39,17 +39,17 @@ func (pop *pipeOffsetProcessor) writeBlock(workerID uint, br *blockResult) {
 	}
 
 	rowsProcessed := pop.rowsProcessed.Add(uint64(len(br.timestamps)))
-	if rowsProcessed <= pop.po.n {
+	if rowsProcessed <= pop.po.offset {
 		return
 	}
 
 	rowsProcessed -= uint64(len(br.timestamps))
-	if rowsProcessed >= pop.po.n {
+	if rowsProcessed >= pop.po.offset {
 		pop.ppBase.writeBlock(workerID, br)
 		return
 	}
 
-	rowsSkip := pop.po.n - rowsProcessed
+	rowsSkip := pop.po.offset - rowsProcessed
 	br.skipRows(int(rowsSkip))
 	pop.ppBase.writeBlock(workerID, br)
 }
@@ -70,7 +70,7 @@ func parsePipeOffset(lex *lexer) (*pipeOffset, error) {
 	}
 	lex.nextToken()
 	po := &pipeOffset{
-		n: n,
+		offset: n,
 	}
 	return po, nil
 }