tsdb

package
v0.3.0 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Feb 6, 2023 License: Apache-2.0 Imports: 36 Imported by: 0

Documentation

Overview

Package tsdb implements a time-series-based storage engine. It provides:

  • Partition data based on a time axis.
  • Sharding data based on a series id which represents a unique entity of stream/measure
  • Retrieving data based on index.Filter.
  • Cleaning expired data, or the data retention.

Index

Constants

This section is empty.

Variables

View Source 
var AnyEntry = Entry(nil)

AnyEntry is the `*` for a regular expression. It could match "any" Entry in an Entity.

View Source 
var (
	// ErrEmptySeriesSpan hints there is no any data blocks based on the input time range.
	ErrEmptySeriesSpan = errors.New("there is no data in such time range")
)

Functions

func GlobalSeriesID 

func GlobalSeriesID(scope Entry) common.SeriesID

GlobalSeriesID encodes Entry to common.SeriesID.

func Hash 

func Hash(entry []byte) []byte

Hash encode Entry to 8 bytes.

func HashEntity 

func HashEntity(entity Entity) []byte

HashEntity runs hash function (e.g. with xxhash algorithm) on each segment of the Entity, and concatenates all uint64 in byte array. So the return length of the byte array will be 8 (every uint64 has 8 bytes) * length of the input.

func MarshalEntityValues added in v0.3.0 

func MarshalEntityValues(evs EntityValues) ([]byte, error)

MarshalEntityValues encodes EntityValues to bytes.

func SeriesID 

func SeriesID(entity Entity) common.SeriesID

SeriesID transforms Entity to common.SeriesID.

Types

type BlockID 

type BlockID struct {
	SegID   SectionID
	BlockID SectionID
}

BlockID is the identity of a block in a shard.

func (BlockID) String added in v0.2.0 

func (b BlockID) String() string

type BlockState 

type BlockState struct {
	TimeRange timestamp.TimeRange
	ID        BlockID
	Closed    bool
}

BlockState is a sample of a block's runtime state.

type CompressionMethod added in v0.3.0 

type CompressionMethod struct {
	Type             CompressionType
	ChunkSizeInBytes int
}

CompressionMethod denotes how to compress a single chunk.

type CompressionType added in v0.3.0 

type CompressionType int

CompressionType specifies how a chunk should be compressed. 

const (
	// CompressionTypeNone mode indicates that a chunk is not compressed.
	CompressionTypeNone CompressionType = iota
	// CompressionTypeZSTD mode indicates that a chunk is compressed using CompressionTypeZSTD algorithm.
	CompressionTypeZSTD
)

type Database 

type Database interface {
	io.Closer
	Shards() []Shard
	Shard(id common.ShardID) (Shard, error)
}

Database allows listing and getting shard details.

func OpenDatabase 

func OpenDatabase(ctx context.Context, opts DatabaseOpts) (Database, error)

OpenDatabase returns a new tsdb runtime. This constructor will create a new database if it's absent, or load an existing one.

type DatabaseOpts 

type DatabaseOpts struct {
	EncodingMethod     EncodingMethod
	Location           string
	CompressionMethod  CompressionMethod
	SegmentInterval    IntervalRule
	BlockInterval      IntervalRule
	TTL                IntervalRule
	BlockMemSize       int64
	BlockInvertedIndex InvertedIndexOpts
	SeriesMemSize      int64
	GlobalIndexMemSize int64
	ShardNum           uint32
	EnableGlobalIndex  bool
}

DatabaseOpts wraps options to create a tsdb.

type EncodingMethod 

type EncodingMethod struct {
	EncoderPool      encoding.SeriesEncoderPool
	DecoderPool      encoding.SeriesDecoderPool
	ChunkSizeInBytes int
}

EncodingMethod wraps encoder/decoder pools to flush/compact data on disk.

type Entity 

type Entity []Entry

Entity denotes an identity of a Series. It defined by Stream or Measure schema.

func NewEntity added in v0.2.0 

func NewEntity(length int) Entity

NewEntity return an Entity with an fixed length.

func (Entity) Copy added in v0.2.0 

func (e Entity) Copy() Entity

Copy an Entity deeply.

func (Entity) Marshal 

func (e Entity) Marshal() []byte

Marshal encodes an Entity to bytes.

func (Entity) Prepend 

func (e Entity) Prepend(entry Entry) Entity

Prepend inserts an Entry before the first Entry as the prefix.

type EntityValue added in v0.3.0 

type EntityValue *modelv1.TagValue

EntityValue represents the value of a tag which is a part of an entity.

func Int64Value added in v0.3.0 

func Int64Value(v int64) EntityValue

Int64Value returns an EntityValue which wraps a int64 value.

func StrValue added in v0.3.0 

func StrValue(v string) EntityValue

StrValue returns an EntityValue which wraps a string value.

type EntityValues added in v0.3.0 

type EntityValues []EntityValue

EntityValues is the encoded Entity.

func DecodeEntityValues added in v0.3.0 

func DecodeEntityValues(tvv []*modelv1.TagValue) (result EntityValues)

DecodeEntityValues decodes tag values to EntityValues.

func UnmarshalEntityValues added in v0.3.0 

func UnmarshalEntityValues(evs []byte) (result EntityValues, err error)

UnmarshalEntityValues decodes EntityValues from bytes.

func (EntityValues) Encode added in v0.3.0 

func (evs EntityValues) Encode() (result []*modelv1.TagValue)

Encode EntityValues to tag values.

func (EntityValues) Prepend added in v0.3.0 

func (evs EntityValues) Prepend(scope EntityValue) EntityValues

Prepend inserts an EntityValue before the first EntityValue as the prefix.

func (EntityValues) String added in v0.3.0 

func (evs EntityValues) String() string

String outputs the string represent of an EntityValue.

func (EntityValues) ToEntity added in v0.3.0 

func (evs EntityValues) ToEntity() (result Entity, err error)

ToEntity transforms EntityValues to Entity.

type Entry 

type Entry []byte

Entry is an element in an Entity.

func EntityValueToEntry added in v0.3.0 

func EntityValueToEntry(ev EntityValue) (Entry, error)

EntityValueToEntry transforms EntityValue to Entry.

type GlobalItemID 

type GlobalItemID struct {
	ShardID common.ShardID

	SeriesID common.SeriesID
	ID       common.ItemID
	// contains filtered or unexported fields
}

GlobalItemID is the top level identity of an item. The item could be retrieved by a GlobalItemID in a tsdb.

type IndexDatabase 

type IndexDatabase interface {
	WriterBuilder() IndexWriterBuilder
	Seek(field index.Field) ([]GlobalItemID, error)
}

IndexDatabase allows stocking index data.

type IndexWriter 

type IndexWriter interface {
	WriteLSMIndex(field []index.Field) error
	WriteInvertedIndex(field []index.Field) error
}

IndexWriter allows ingesting index data.

type IndexWriterBuilder 

type IndexWriterBuilder interface {
	Scope(scope Entry) IndexWriterBuilder
	Time(ts time.Time) IndexWriterBuilder
	GlobalItemID(itemID GlobalItemID) IndexWriterBuilder
	Build() (IndexWriter, error)
}

IndexWriterBuilder is a helper to build IndexWriter.

type IntervalRule 

type IntervalRule struct {
	Unit IntervalUnit
	Num  int
}

IntervalRule defines a length of two points in time.

type IntervalUnit 

type IntervalUnit int

IntervalUnit denotes the unit of a time point. 

const (
	HOUR IntervalUnit = iota
	DAY
)

Available IntervalUnits. HOUR and DAY are adequate for the APM scenario.

func (IntervalUnit) String 

func (iu IntervalUnit) String() string

type InvertedIndexOpts added in v0.3.0 

type InvertedIndexOpts struct {
	BatchWaitSec int64
}

InvertedIndexOpts wraps options to create the block inverted index.

type Item 

type Item interface {
	Family(family []byte) ([]byte, error)
	Val() ([]byte, error)
	ID() common.ItemID
	SortedField() []byte
	Time() uint64
}

Item allows retrieving raw data from an item.

type Iterator 

type Iterator interface {
	Next() bool
	Val() Item
	Close() error
}

Iterator allows iterating a series in a time span.

type Path 

type Path struct {
	// contains filtered or unexported fields
}

Path denotes a expression to match a Series. It supports the fuzzy matching more than EQ by setting an entry to AnyEntry.

func NewPath 

func NewPath(matchingExpression []Entry) Path

NewPath return a Path with a matching expression.

type SectionID added in v0.3.0 

type SectionID uint32

SectionID is the kind of a block/segment.

func GenerateInternalID 

func GenerateInternalID(unit IntervalUnit, suffix int) SectionID

GenerateInternalID returns a identity of a section(segment or block) based on IntervalRule.

type Seeker 

type Seeker interface {
	Seek() ([]Iterator, error)
}

Seeker allows searching data in a Database.

type SeekerBuilder 

type SeekerBuilder interface {
	Filter(predicator index.Filter) SeekerBuilder
	OrderByIndex(indexRule *databasev1.IndexRule, order modelv1.Sort) SeekerBuilder
	OrderByTime(order modelv1.Sort) SeekerBuilder
	Build() (Seeker, error)
}

SeekerBuilder a helper to build a Seeker.

type Series 

type Series interface {
	ID() common.SeriesID
	Span(ctx context.Context, timeRange timestamp.TimeRange) (SeriesSpan, error)
	Create(ctx context.Context, t time.Time) (SeriesSpan, error)
	Get(ctx context.Context, id GlobalItemID) (Item, io.Closer, error)
	String() string
}

Series denotes a series of data points group by a common.SeriesID common.SeriesID is encoded by a entity defined by Stream or Measure.

type SeriesDatabase 

type SeriesDatabase interface {
	observability.Observable
	io.Closer
	GetByID(id common.SeriesID) (Series, error)
	Get(key []byte, entityValues EntityValues) (Series, error)
	List(ctx context.Context, path Path) (SeriesList, error)
}

SeriesDatabase allows retrieving series.

type SeriesList 

type SeriesList []Series

SeriesList is a collection of Series.

func (SeriesList) Len 

func (a SeriesList) Len() int

func (SeriesList) Less 

func (a SeriesList) Less(i, j int) bool

func (SeriesList) Merge added in v0.2.0 

func (a SeriesList) Merge(other SeriesList) SeriesList

Merge other SeriesList with this one to create a new SeriesList.

func (SeriesList) Swap 

func (a SeriesList) Swap(i, j int)

type SeriesSpan 

type SeriesSpan interface {
	io.Closer
	WriterBuilder() WriterBuilder
	SeekerBuilder() SeekerBuilder
}

SeriesSpan is a span in a time series. It contains data blocks in such time range.

type Shard 

type Shard interface {
	io.Closer
	ID() common.ShardID
	Series() SeriesDatabase
	Index() IndexDatabase
	State() ShardState
	// Only works with MockClock
	TriggerSchedule(task string) bool
}

Shard allows accessing data of tsdb.

func NewScopedShard 

func NewScopedShard(scope Entry, delegated Shard) Shard

NewScopedShard returns a shard in a scope.

func OpenShard 

func OpenShard(ctx context.Context, id common.ShardID,
	root string, segmentSize, blockSize, ttl IntervalRule, openedBlockSize, maxOpenedBlockSize int,
) (Shard, error)

OpenShard returns an existed Shard or create a new one if not existed.

type ShardState 

type ShardState struct {
	Blocks           []BlockState
	OpenBlocks       []BlockID
	StrategyManagers []string
}

ShardState is a sample of a shard's runtime state.

type Supplier 

type Supplier interface {
	SupplyTSDB() Database
}

Supplier allows getting a tsdb's runtime.

type Writer 

type Writer interface {
	IndexWriter
	Write() (GlobalItemID, error)
	ItemID() GlobalItemID
	String() string
}

Writer allow ingesting data into a tsdb.

type WriterBuilder 

type WriterBuilder interface {
	Family(name []byte, val []byte) WriterBuilder
	Time(ts time.Time) WriterBuilder
	Val(val []byte) WriterBuilder
	Build() (Writer, error)
}

WriterBuilder is a helper to build a Writer.

Source Files

View all Source files

Directories

Path Synopsis
Package bucket implements a rolling bucket system.
 Package bucket implements a rolling bucket system.
Package index implements transferring data to indices.
 Package index implements transferring data to indices.

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.