prometheus/writer.go

494 lines
12 KiB
Go
Raw Normal View History

2016-12-09 19:45:46 +00:00
package tsdb
import (
2016-12-09 20:23:34 +00:00
"encoding/binary"
"fmt"
2016-12-09 19:45:46 +00:00
"hash/crc32"
"io"
2016-12-09 20:23:34 +00:00
"sort"
"strings"
"github.com/bradfitz/slice"
2017-01-01 19:01:17 +00:00
"github.com/fabxc/tsdb/chunks"
2016-12-21 08:39:01 +00:00
"github.com/fabxc/tsdb/labels"
"github.com/pkg/errors"
2016-12-09 19:45:46 +00:00
)
const (
// MagicSeries 4 bytes at the head of series file.
MagicSeries = 0x85BD40DD
// MagicIndex 4 bytes at the head of an index file.
MagicIndex = 0xBAAAD700
)
// SeriesWriter serializes a time block of chunked series data.
type SeriesWriter interface {
// WriteSeries writes the time series data chunks for a single series.
2016-12-10 08:44:00 +00:00
// The reference is used to resolve the correct series in the written index.
// It only has to be valid for the duration of the write.
2017-01-01 19:01:17 +00:00
WriteSeries(ref uint32, l labels.Labels, chunks []ChunkMeta) error
2016-12-09 19:45:46 +00:00
// Size returns the size of the data written so far.
Size() int64
// Close writes any required finalization and closes the resources
// associated with the underlying writer.
Close() error
}
// seriesWriter implements the SeriesWriter interface for the standard
// serialization format.
type seriesWriter struct {
w io.Writer
n int64
c int
2017-01-01 19:01:17 +00:00
index IndexWriter
2016-12-09 19:45:46 +00:00
}
2017-01-01 19:01:17 +00:00
func newSeriesWriter(w io.Writer, index IndexWriter) *seriesWriter {
2016-12-09 19:45:46 +00:00
return &seriesWriter{
2017-01-01 19:01:17 +00:00
w: w,
n: 0,
index: index,
2016-12-09 19:45:46 +00:00
}
}
func (w *seriesWriter) write(wr io.Writer, b []byte) error {
n, err := wr.Write(b)
w.n += int64(n)
return err
}
func (w *seriesWriter) writeMeta() error {
2016-12-11 14:49:24 +00:00
b := [8]byte{}
2016-12-09 19:45:46 +00:00
binary.BigEndian.PutUint32(b[:4], MagicSeries)
b[4] = flagStd
return w.write(w.w, b[:])
2016-12-09 19:45:46 +00:00
}
2017-01-01 19:01:17 +00:00
func (w *seriesWriter) WriteSeries(ref uint32, lset labels.Labels, chks []ChunkMeta) error {
2016-12-09 19:45:46 +00:00
// Initialize with meta data.
if w.n == 0 {
if err := w.writeMeta(); err != nil {
return err
}
}
// TODO(fabxc): is crc32 enough for chunks of one series?
h := crc32.NewIEEE()
wr := io.MultiWriter(h, w.w)
2016-12-11 14:49:24 +00:00
// For normal reads we don't need the number of the chunk section but
2016-12-09 19:45:46 +00:00
// it allows us to verify checksums without reading the index file.
2016-12-11 14:49:24 +00:00
// The offsets are also technically enough to calculate chunk size. but
// holding the length of each chunk could later allow for adding padding
// between chunks.
b := [binary.MaxVarintLen32]byte{}
n := binary.PutUvarint(b[:], uint64(len(chks)))
2016-12-11 14:49:24 +00:00
if err := w.write(wr, b[:n]); err != nil {
2016-12-09 19:45:46 +00:00
return err
}
2017-01-01 19:01:17 +00:00
for i := range chks {
chk := &chks[i]
2016-12-09 19:45:46 +00:00
2017-01-01 19:01:17 +00:00
chk.Ref = uint32(w.n)
n = binary.PutUvarint(b[:], uint64(len(chk.Chunk.Bytes())))
2016-12-09 19:45:46 +00:00
2016-12-11 14:49:24 +00:00
if err := w.write(wr, b[:n]); err != nil {
return err
}
2017-01-01 19:01:17 +00:00
if err := w.write(wr, []byte{byte(chk.Chunk.Encoding())}); err != nil {
2016-12-09 19:45:46 +00:00
return err
}
2017-01-01 19:01:17 +00:00
if err := w.write(wr, chk.Chunk.Bytes()); err != nil {
2016-12-09 19:45:46 +00:00
return err
}
2017-01-01 19:01:17 +00:00
chk.Chunk = nil
2016-12-09 19:45:46 +00:00
}
if err := w.write(w.w, h.Sum(nil)); err != nil {
return err
}
if w.index != nil {
2017-01-01 19:01:17 +00:00
w.index.AddSeries(ref, lset, chks...)
2016-12-09 19:45:46 +00:00
}
return nil
}
func (w *seriesWriter) Size() int64 {
return w.n
}
func (w *seriesWriter) Close() error {
return nil
}
2017-01-01 19:01:17 +00:00
// ChunkMeta holds information about a chunk of data.
type ChunkMeta struct {
2017-01-01 19:01:17 +00:00
// Ref and Chunk hold either a reference that can be used to retrieve
// chunk data or the data itself.
// Generally, only one of them is set.
Ref uint32
Chunk chunks.Chunk
MinTime, MaxTime int64 // time range the data covers
2016-12-09 19:45:46 +00:00
}
// IndexWriter serialized the index for a block of series data.
// The methods must generally be called in order they are specified.
type IndexWriter interface {
2016-12-10 08:44:00 +00:00
// AddSeries populates the index writer witha series and its offsets
// of chunks that the index can reference.
// The reference number is used to resolve a series against the postings
// list iterator. It only has to be available during the write processing.
2016-12-21 08:39:01 +00:00
AddSeries(ref uint32, l labels.Labels, chunks ...ChunkMeta)
2016-12-09 19:45:46 +00:00
// WriteStats writes final stats for the indexed block.
WriteStats(BlockStats) error
2016-12-09 19:45:46 +00:00
// WriteLabelIndex serializes an index from label names to values.
// The passed in values chained tuples of strings of the length of names.
WriteLabelIndex(names []string, values []string) error
// WritePostings writes a postings list for a single label pair.
2016-12-14 20:58:29 +00:00
WritePostings(name, value string, it Postings) error
2016-12-09 19:45:46 +00:00
// Size returns the size of the data written so far.
Size() int64
2016-12-09 20:23:34 +00:00
// Close writes any finalization and closes theresources associated with
2016-12-09 19:45:46 +00:00
// the underlying writer.
Close() error
}
2016-12-10 08:44:00 +00:00
type indexWriterSeries struct {
2016-12-21 08:39:01 +00:00
labels labels.Labels
chunks []ChunkMeta // series file offset of chunks
offset uint32 // index file offset of series reference
2016-12-10 08:44:00 +00:00
}
2016-12-09 19:45:46 +00:00
// indexWriter implements the IndexWriter interface for the standard
// serialization format.
type indexWriter struct {
w io.Writer
n int64
2016-12-10 08:44:00 +00:00
series map[uint32]*indexWriterSeries
2016-12-09 20:23:34 +00:00
2016-12-09 20:40:38 +00:00
symbols map[string]uint32 // symbol offsets
labelIndexes []hashEntry // label index offsets
2016-12-10 08:44:00 +00:00
postings []hashEntry // postings lists offsets
2016-12-09 20:23:34 +00:00
}
func newIndexWriter(w io.Writer) *indexWriter {
return &indexWriter{
w: w,
n: 0,
symbols: make(map[string]uint32, 4096),
series: make(map[uint32]*indexWriterSeries, 4096),
2016-12-09 20:23:34 +00:00
}
}
func (w *indexWriter) write(wr io.Writer, b []byte) error {
n, err := wr.Write(b)
w.n += int64(n)
return err
}
2016-12-09 21:36:31 +00:00
// section writes a CRC32 checksummed section of length l and guarded by flag.
func (w *indexWriter) section(l uint32, flag byte, f func(w io.Writer) error) error {
h := crc32.NewIEEE()
wr := io.MultiWriter(h, w.w)
b := [5]byte{flag, 0, 0, 0, 0}
binary.BigEndian.PutUint32(b[1:], l)
if err := w.write(wr, b[:]); err != nil {
return errors.Wrap(err, "writing header")
2016-12-09 21:36:31 +00:00
}
if err := f(wr); err != nil {
return errors.Wrap(err, "contents write func")
}
if err := w.write(w.w, h.Sum(nil)); err != nil {
return errors.Wrap(err, "writing checksum")
2016-12-09 21:36:31 +00:00
}
return nil
2016-12-09 21:36:31 +00:00
}
2016-12-09 20:23:34 +00:00
func (w *indexWriter) writeMeta() error {
2016-12-11 14:49:24 +00:00
b := [8]byte{}
binary.BigEndian.PutUint32(b[:4], MagicIndex)
b[4] = flagStd
2016-12-09 20:23:34 +00:00
return w.write(w.w, b[:])
2016-12-09 19:45:46 +00:00
}
2016-12-21 08:39:01 +00:00
func (w *indexWriter) AddSeries(ref uint32, lset labels.Labels, chunks ...ChunkMeta) {
2016-12-09 20:23:34 +00:00
// Populate the symbol table from all label sets we have to reference.
for _, l := range lset {
w.symbols[l.Name] = 0
w.symbols[l.Value] = 0
}
2016-12-10 08:44:00 +00:00
w.series[ref] = &indexWriterSeries{
labels: lset,
chunks: chunks,
2016-12-10 08:44:00 +00:00
}
2016-12-09 19:45:46 +00:00
}
func (w *indexWriter) WriteStats(stats BlockStats) error {
if w.n != 0 {
return fmt.Errorf("WriteStats must be called first")
}
if err := w.writeMeta(); err != nil {
return err
}
b := [64]byte{}
binary.BigEndian.PutUint64(b[0:], uint64(stats.MinTime))
binary.BigEndian.PutUint64(b[8:], uint64(stats.MaxTime))
binary.BigEndian.PutUint32(b[16:], stats.SeriesCount)
binary.BigEndian.PutUint32(b[20:], stats.ChunkCount)
binary.BigEndian.PutUint64(b[24:], stats.SampleCount)
err := w.section(64, flagStd, func(wr io.Writer) error {
return w.write(wr, b[:])
})
if err != nil {
return err
}
if err := w.writeSymbols(); err != nil {
return err
}
if err := w.writeSeries(); err != nil {
return err
2016-12-09 20:23:34 +00:00
}
2016-12-09 19:45:46 +00:00
return nil
}
2016-12-09 20:23:34 +00:00
func (w *indexWriter) writeSymbols() error {
// Generate sorted list of strings we will store as reference table.
symbols := make([]string, 0, len(w.symbols))
for s := range w.symbols {
symbols = append(symbols, s)
}
sort.Strings(symbols)
// The start of the section plus a 5 byte section header are our base.
// TODO(fabxc): switch to relative offsets and hold sections in a TOC.
base := uint32(w.n) + 5
2016-12-09 21:27:43 +00:00
buf := [binary.MaxVarintLen32]byte{}
2016-12-11 14:50:00 +00:00
b := append(make([]byte, 0, 4096), flagStd)
2016-12-09 20:23:34 +00:00
for _, s := range symbols {
w.symbols[s] = base + uint32(len(b))
2016-12-09 21:27:43 +00:00
n := binary.PutUvarint(buf[:], uint64(len(s)))
b = append(b, buf[:n]...)
b = append(b, s...)
}
2016-12-09 20:23:34 +00:00
l := uint32(len(b))
2016-12-09 21:27:43 +00:00
2016-12-09 21:36:31 +00:00
return w.section(l, flagStd, func(wr io.Writer) error {
return w.write(wr, b)
})
2016-12-09 21:27:43 +00:00
}
func (w *indexWriter) writeSeries() error {
// Series must be stored sorted along their labels.
series := make([]*indexWriterSeries, 0, len(w.series))
2016-12-09 21:27:43 +00:00
2016-12-10 08:44:00 +00:00
for _, s := range w.series {
series = append(series, s)
}
slice.Sort(series, func(i, j int) bool {
2016-12-24 12:49:35 +00:00
return labels.Compare(series[i].labels, series[j].labels) < 0
})
// Current end of file plus 5 bytes for section header.
// TODO(fabxc): switch to relative offsets.
base := uint32(w.n) + 5
b := make([]byte, 0, 1<<20) // 1MiB
buf := make([]byte, binary.MaxVarintLen64)
for _, s := range series {
// Write label set symbol references.
s.offset = base + uint32(len(b))
2016-12-10 08:44:00 +00:00
n := binary.PutUvarint(buf, uint64(len(s.labels)))
2016-12-10 08:44:00 +00:00
b = append(b, buf[:n]...)
2016-12-09 21:27:43 +00:00
2016-12-10 08:44:00 +00:00
for _, l := range s.labels {
n = binary.PutUvarint(buf, uint64(w.symbols[l.Name]))
2016-12-09 21:27:43 +00:00
b = append(b, buf[:n]...)
n = binary.PutUvarint(buf, uint64(w.symbols[l.Value]))
b = append(b, buf[:n]...)
}
// Write chunks meta data including reference into chunk file.
n = binary.PutUvarint(buf, uint64(len(s.chunks)))
b = append(b, buf[:n]...)
for _, c := range s.chunks {
n = binary.PutVarint(buf, c.MinTime)
b = append(b, buf[:n]...)
n = binary.PutVarint(buf, c.MaxTime)
b = append(b, buf[:n]...)
n = binary.PutUvarint(buf, uint64(c.Ref))
2016-12-09 21:27:43 +00:00
b = append(b, buf[:n]...)
}
}
l := uint32(len(b))
2016-12-09 21:27:43 +00:00
2016-12-09 21:36:31 +00:00
return w.section(l, flagStd, func(wr io.Writer) error {
return w.write(wr, b)
})
2016-12-09 19:45:46 +00:00
}
func (w *indexWriter) WriteLabelIndex(names []string, values []string) error {
valt, err := newStringTuples(values, len(names))
if err != nil {
return err
2016-12-09 20:40:38 +00:00
}
sort.Sort(valt)
2016-12-09 20:40:38 +00:00
w.labelIndexes = append(w.labelIndexes, hashEntry{
name: strings.Join(names, string(sep)),
offset: uint32(w.n),
})
buf := make([]byte, binary.MaxVarintLen32)
n := binary.PutUvarint(buf, uint64(len(names)))
l := uint32(n) + uint32(len(values)*4)
2016-12-09 21:36:31 +00:00
return w.section(l, flagStd, func(wr io.Writer) error {
// First byte indicates tuple size for index.
if err := w.write(wr, buf[:n]); err != nil {
return err
}
for _, v := range valt.s {
binary.BigEndian.PutUint32(buf, w.symbols[v])
2016-12-09 20:40:38 +00:00
if err := w.write(wr, buf[:4]); err != nil {
2016-12-09 21:36:31 +00:00
return err
}
2016-12-09 20:40:38 +00:00
}
2016-12-09 21:36:31 +00:00
return nil
})
2016-12-09 19:45:46 +00:00
}
2016-12-14 20:58:29 +00:00
func (w *indexWriter) WritePostings(name, value string, it Postings) error {
2016-12-10 08:44:00 +00:00
key := name + string(sep) + value
w.postings = append(w.postings, hashEntry{
name: key,
offset: uint32(w.n),
})
b := make([]byte, 0, 4096)
buf := [4]byte{}
2016-12-10 08:44:00 +00:00
// Order of the references in the postings list does not imply order
// of the series references within the persisted block they are mapped to.
// We have to sort the new references again.
var refs []uint32
2016-12-10 08:44:00 +00:00
for it.Next() {
2017-01-02 12:27:52 +00:00
s, ok := w.series[it.At()]
if !ok {
2017-01-02 12:27:52 +00:00
return errors.Errorf("series for reference %d not found", it.At())
}
refs = append(refs, s.offset)
}
if err := it.Err(); err != nil {
return err
}
slice.Sort(refs, func(i, j int) bool { return refs[i] < refs[j] })
for _, r := range refs {
binary.BigEndian.PutUint32(buf[:], r)
b = append(b, buf[:]...)
2016-12-10 08:44:00 +00:00
}
return w.section(uint32(len(b)), flagStd, func(wr io.Writer) error {
return w.write(wr, b)
})
2016-12-09 19:45:46 +00:00
}
func (w *indexWriter) Size() int64 {
return w.n
}
2016-12-09 20:23:34 +00:00
type hashEntry struct {
name string
offset uint32
}
func (w *indexWriter) writeHashmap(h []hashEntry) error {
b := make([]byte, 0, 4096)
buf := [binary.MaxVarintLen32]byte{}
for _, e := range h {
n := binary.PutUvarint(buf[:], uint64(len(e.name)))
b = append(b, buf[:n]...)
b = append(b, e.name...)
n = binary.PutUvarint(buf[:], uint64(e.offset))
b = append(b, buf[:n]...)
}
return w.section(uint32(len(b)), flagStd, func(wr io.Writer) error {
return w.write(wr, b)
2016-12-09 21:36:31 +00:00
})
}
2016-12-10 08:44:00 +00:00
func (w *indexWriter) finalize() error {
// Write out hash maps to jump to correct label index and postings sections.
lo := uint32(w.n)
if err := w.writeHashmap(w.labelIndexes); err != nil {
2016-12-10 08:44:00 +00:00
return err
}
po := uint32(w.n)
if err := w.writeHashmap(w.postings); err != nil {
return err
}
2016-12-10 08:44:00 +00:00
// Terminate index file with offsets to hashmaps. This is the entry Pointer
// for any index query.
// TODO(fabxc): also store offset to series section to allow plain
// iteration over all existing series?
// TODO(fabxc): store references like these that are not resolved via direct
// mmap using explicit endianness?
b := [8]byte{}
binary.BigEndian.PutUint32(b[:4], lo)
binary.BigEndian.PutUint32(b[4:], po)
2016-12-10 08:44:00 +00:00
return w.write(w.w, b[:])
}
2016-12-09 19:45:46 +00:00
func (w *indexWriter) Close() error {
2017-01-02 13:41:13 +00:00
return w.finalize()
2016-12-09 19:45:46 +00:00
}