tsdb

package
v1.7.2 Latest Latest
Warning

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

 Go to latest Published: Dec 12, 2018 License: MIT Imports: 42 Imported by: 0

README

Line Protocol

The line protocol is a text based format for writing points to InfluxDB. Each line defines a single point. Multiple lines must be separated by the newline character \n. The format of the line consists of three parts:

[key] [fields] [timestamp]

Each section is separated by spaces. The minimum required point consists of a measurement name and at least one field. Points without a specified timestamp will be written using the server's local timestamp. Timestamps are assumed to be in nanoseconds unless a precision value is passed in the query string.

Key

The key is the measurement name and any optional tags separated by commas. Measurement names, tag keys, and tag values must escape any spaces or commas using a backslash (\). For example: \ and \,. All tag values are stored as strings and should not be surrounded in quotes.

Tags should be sorted by key before being sent for best performance. The sort should match that from the Go bytes.Compare function (http://golang.org/pkg/bytes/#Compare).

Examples
# measurement only
cpu

# measurement and tags
cpu,host=serverA,region=us-west

# measurement with commas
cpu\,01,host=serverA,region=us-west

# tag value with spaces
cpu,host=server\ A,region=us\ west

Fields

Fields are key-value metrics associated with the measurement. Every line must have at least one field. Multiple fields must be separated with commas and not spaces.

Field keys are always strings and follow the same syntactical rules as described above for tag keys and values. Field values can be one of four types. The first value written for a given field on a given measurement defines the type of that field for all series under that measurement.

  • integer - Numeric values that do not include a decimal and are followed by a trailing i when inserted (e.g. 1i, 345i, 2015i, -10i). Note that all values must have a trailing i. If they do not they will be written as floats.
  • float - Numeric values that are not followed by a trailing i. (e.g. 1, 1.0, -3.14, 6.0+e5, 10).
  • boolean - A value indicating true or false. Valid boolean strings are (t, T, true, TRUE, f, F, false, and FALSE).
  • string - A text value. All string values must be surrounded in double-quotes ". If the string contains a double-quote or backslashes, it must be escaped with a backslash, e.g. \", \\.
# integer value
cpu value=1i

cpu value=1.1i # will result in a parse error

# float value
cpu_load value=1

cpu_load value=1.0

cpu_load value=1.2

# boolean value
error fatal=true

# string value
event msg="logged out"

# multiple values
cpu load=10,alert=true,reason="value above maximum threshold"

Timestamp

The timestamp section is optional but should be specified if possible. The value is an integer representing nanoseconds since the epoch. If the timestamp is not provided the point will inherit the server's local timestamp.

Some write APIs allow passing a lower precision. If the API supports a lower precision, the timestamp may also be an integer epoch in microseconds, milliseconds, seconds, minutes or hours.

Full Example

A full example is shown below.

cpu,host=server01,region=uswest value=1 1434055562000000000
cpu,host=server02,region=uswest value=3 1434055562000010000

In this example the first line shows a measurement of "cpu", there are two tags "host" and "region, the value is 1.0, and the timestamp is 1434055562000000000. Following this is a second line, also a point in the measurement "cpu" but belonging to a different "host".

cpu,host=server\ 01,region=uswest value=1,msg="all systems nominal"
cpu,host=server\ 01,region=us\,west value_int=1i

In these examples, the "host" is set to server 01. The field value associated with field key msg is double-quoted, as it is a string. The second example shows a region of us,west with the comma properly escaped. In the first example value is written as a floating point number. In the second, value_int is an integer.

Distributed Queries

Documentation

Index

Constants

View Source 
const (
	// DefaultEngine is the default engine for new shards
	DefaultEngine = "tsm1"

	// DefaultIndex is the default index for new shards
	DefaultIndex = InmemIndexName

	// DefaultCacheMaxMemorySize is the maximum size a shard's cache can
	// reach before it starts rejecting writes.
	DefaultCacheMaxMemorySize = 1024 * 1024 * 1024 // 1GB

	// DefaultCacheSnapshotMemorySize is the size at which the engine will
	// snapshot the cache and write it to a TSM file, freeing up memory
	DefaultCacheSnapshotMemorySize = 25 * 1024 * 1024 // 25MB

	// DefaultCacheSnapshotWriteColdDuration is the length of time at which
	// the engine will snapshot the cache and write it to a new TSM file if
	// the shard hasn't received writes or deletes
	DefaultCacheSnapshotWriteColdDuration = time.Duration(10 * time.Minute)

	// DefaultCompactFullWriteColdDuration is the duration at which the engine
	// will compact all TSM files in a shard if it hasn't received a write or delete
	DefaultCompactFullWriteColdDuration = time.Duration(4 * time.Hour)

	// DefaultCompactThroughput is the rate limit in bytes per second that we
	// will allow TSM compactions to write to disk. Not that short bursts are allowed
	// to happen at a possibly larger value, set by DefaultCompactThroughputBurst.
	// A value of 0 here will disable compaction rate limiting
	DefaultCompactThroughput = 48 * 1024 * 1024

	// DefaultCompactThroughputBurst is the rate limit in bytes per second that we
	// will allow TSM compactions to write to disk. If this is not set, the burst value
	// will be set to equal the normal throughput
	DefaultCompactThroughputBurst = 48 * 1024 * 1024

	// DefaultMaxPointsPerBlock is the maximum number of points in an encoded
	// block in a TSM file
	DefaultMaxPointsPerBlock = 1000

	// DefaultMaxSeriesPerDatabase is the maximum number of series a node can hold per database.
	// This limit only applies to the "inmem" index.
	DefaultMaxSeriesPerDatabase = 1000000

	// DefaultMaxValuesPerTag is the maximum number of values a tag can have within a measurement.
	DefaultMaxValuesPerTag = 100000

	// DefaultMaxConcurrentCompactions is the maximum number of concurrent full and level compactions
	// that can run at one time.  A value of 0 results in 50% of runtime.GOMAXPROCS(0) used at runtime.
	DefaultMaxConcurrentCompactions = 0

	// DefaultMaxIndexLogFileSize is the default threshold, in bytes, when an index
	// write-ahead log file will compact into an index file.
	DefaultMaxIndexLogFileSize = 1 * 1024 * 1024 // 1MB
)
View Source 
const (
	InmemIndexName = "inmem"
	TSI1IndexName  = "tsi1"
)

Available index types.

View Source 
const (
	SeriesIndexVersion = 1
	SeriesIndexMagic   = "SIDX"
)
View Source 
const (
	SeriesIndexElemSize   = 16 // offset + id
	SeriesIndexLoadFactor = 90 // rhh load factor

	SeriesIndexHeaderSize = 0 +
		4 + 1 +
		8 + 8 +
		8 + 8 +
		8 + 8 +
		8 + 8 +
		0
)
View Source 
const (
	SeriesSegmentVersion = 1
	SeriesSegmentMagic   = "SSEG"

	SeriesSegmentHeaderSize = 4 + 1 // magic + version
)
View Source 
const (
	SeriesEntryFlagSize   = 1
	SeriesEntryHeaderSize = 1 + 8 // flag + id

	SeriesEntryInsertFlag    = 0x01
	SeriesEntryTombstoneFlag = 0x02
)

Series entry constants.

View Source 
const DefaultSeriesPartitionCompactThreshold = 1 << 17 // 128K

DefaultSeriesPartitionCompactThreshold is the number of series IDs to hold in the in-memory series map before compacting and rebuilding the on-disk representation.

View Source 
const EOF = query.ZeroTime

EOF represents a "not found" key returned by a Cursor.

View Source 
const SeriesFileDirectory = "_series"

SeriesFileDirectory is the name of the directory containing series files for a database.

View Source 
const (
	// SeriesFilePartitionN is the number of partitions a series file is split into.
	SeriesFilePartitionN = 8
)
View Source 
const SeriesIDSize = 8

SeriesIDSize is the size in bytes of a series key ID.

Variables

View Source 
var (
	// ErrFormatNotFound is returned when no format can be determined from a path.
	ErrFormatNotFound = errors.New("format not found")

	// ErrUnknownEngineFormat is returned when the engine format is
	// unknown. ErrUnknownEngineFormat is currently returned if a format
	// other than tsm1 is encountered.
	ErrUnknownEngineFormat = errors.New("unknown engine format")
)
View Source 
var (
	ErrSeriesFileClosed         = errors.New("tsdb: series file closed")
	ErrInvalidSeriesPartitionID = errors.New("tsdb: invalid series partition id")
)
View Source 
var (
	ErrSeriesPartitionClosed              = errors.New("tsdb: series partition closed")
	ErrSeriesPartitionCompactionCancelled = errors.New("tsdb: series partition compaction cancelled")
)
View Source 
var (
	ErrInvalidSeriesSegment        = errors.New("invalid series segment")
	ErrInvalidSeriesSegmentVersion = errors.New("invalid series segment version")
	ErrSeriesSegmentNotWritable    = errors.New("series segment not writable")
)
View Source 
var (
	// ErrFieldOverflow is returned when too many fields are created on a measurement.
	ErrFieldOverflow = errors.New("field overflow")

	// ErrFieldTypeConflict is returned when a new field already exists with a different type.
	ErrFieldTypeConflict = errors.New("field type conflict")

	// ErrFieldNotFound is returned when a field cannot be found.
	ErrFieldNotFound = errors.New("field not found")

	// ErrFieldUnmappedID is returned when the system is presented, during decode, with a field ID
	// there is no mapping for.
	ErrFieldUnmappedID = errors.New("field ID not mapped")

	// ErrEngineClosed is returned when a caller attempts indirectly to
	// access the shard's underlying engine.
	ErrEngineClosed = errors.New("engine is closed")

	// ErrShardDisabled is returned when a the shard is not available for
	// queries or writes.
	ErrShardDisabled = errors.New("shard is disabled")

	// ErrUnknownFieldsFormat is returned when the fields index file is not identifiable by
	// the file's magic number.
	ErrUnknownFieldsFormat = errors.New("unknown field index format")

	// ErrUnknownFieldType is returned when the type of a field cannot be determined.
	ErrUnknownFieldType = errors.New("unknown field type")

	// ErrShardNotIdle is returned when an operation requring the shard to be idle/cold is
	// attempted on a hot shard.
	ErrShardNotIdle = errors.New("shard not idle")
)
View Source 
var (
	// ErrShardNotFound is returned when trying to get a non existing shard.
	ErrShardNotFound = fmt.Errorf("shard not found")
	// ErrStoreClosed is returned when trying to use a closed Store.
	ErrStoreClosed = fmt.Errorf("store is closed")
	// ErrShardDeletion is returned when trying to create a shard that is being deleted
	ErrShardDeletion = errors.New("shard is being deleted")
	// ErrMultipleIndexTypes is returned when trying to do deletes on a database with
	// multiple index types.
	ErrMultipleIndexTypes = errors.New("cannot delete data. DB contains shards using both inmem and tsi1 indexes. Please convert all shards to use the same index type to delete data.")
)
View Source 
var ErrIndexClosing = errors.New("index is closing")

ErrIndexClosing can be returned to from an Index method if the index is currently closing.

View Source 
var ErrInvalidSeriesIndex = errors.New("invalid series index")
View Source 
var NewInmemIndex func(name string, sfile *SeriesFile) (interface{}, error)

NewInmemIndex returns a new "inmem" index type.

Functions

func AppendSeriesEntry added in v1.5.0 

func AppendSeriesEntry(dst []byte, flag uint8, id uint64, key []byte) []byte

func AppendSeriesKey added in v1.5.0 

func AppendSeriesKey(dst []byte, name []byte, tags models.Tags) []byte

AppendSeriesKey serializes name and tags to a byte slice. The total length is prepended as a uvarint.

func CompareSeriesKeys added in v1.5.0 

func CompareSeriesKeys(a, b []byte) int

func GenerateSeriesKeys added in v1.5.0 

func GenerateSeriesKeys(names [][]byte, tagsSlice []models.Tags) [][]byte

GenerateSeriesKeys generates series keys for a list of names & tags using a single large memory block.

func IsValidSeriesEntryFlag added in v1.6.1 

func IsValidSeriesEntryFlag(flag byte) bool

IsValidSeriesEntryFlag returns true if flag is valid.

func IsValidSeriesSegmentFilename added in v1.5.0 

func IsValidSeriesSegmentFilename(filename string) bool

IsValidSeriesSegmentFilename returns true if filename is a 4-character lowercase hexidecimal number.

func JoinSeriesOffset added in v1.5.0 

func JoinSeriesOffset(segmentID uint16, pos uint32) int64

JoinSeriesOffset returns an offset that combines the 2-byte segmentID and 4-byte pos.

func MakeTagsKey added in v1.3.0 

func MakeTagsKey(keys []string, tags models.Tags) []byte

MakeTagsKey converts a tag set to bytes for use as a lookup key.

func MarshalTags added in v0.9.3 

func MarshalTags(tags map[string]string) []byte

MarshalTags converts a tag set to bytes for use as a lookup key.

func NewFieldKeysIterator added in v0.11.0 

func NewFieldKeysIterator(sh *Shard, opt query.IteratorOptions) (query.Iterator, error)

NewFieldKeysIterator returns an iterator that can be iterated over to retrieve field keys.

func NewMeasurementSliceIterator added in v1.5.0 

func NewMeasurementSliceIterator(names [][]byte) *measurementSliceIterator

NewMeasurementSliceIterator returns an iterator over a slice of in-memory measurement names.

func NewSeriesPointIterator added in v1.5.0 

func NewSeriesPointIterator(indexSet IndexSet, opt query.IteratorOptions) (_ query.Iterator, err error)

newSeriesPointIterator returns a new instance of seriesPointIterator.

func NewSeriesQueryAdapterIterator added in v1.5.0 

func NewSeriesQueryAdapterIterator(sfile *SeriesFile, itr SeriesIDIterator, fieldset *MeasurementFieldSet, opt query.IteratorOptions) query.Iterator

NewSeriesQueryAdapterIterator returns a new instance of SeriesQueryAdapterIterator.

func NewShardError added in v0.11.0 

func NewShardError(id uint64, err error) error

NewShardError returns a new ShardError.

func NewTagKeySliceIterator added in v1.5.0 

func NewTagKeySliceIterator(keys [][]byte) *tagKeySliceIterator

NewTagKeySliceIterator returns a TagKeyIterator that iterates over a slice.

func NewTagKeysIterator added in v0.11.0 

func NewTagKeysIterator(sh *Shard, opt query.IteratorOptions) (query.Iterator, error)

NewTagKeysIterator returns a new instance of TagKeysIterator.

func NewTagValueSliceIterator added in v1.5.0 

func NewTagValueSliceIterator(values [][]byte) *tagValueSliceIterator

NewTagValueSliceIterator returns a TagValueIterator that iterates over a slice.

func ParseSeriesKey added in v1.5.0 

func ParseSeriesKey(data []byte) (name []byte, tags models.Tags)

ParseSeriesKey extracts the name & tags from a series key.

func ParseSeriesKeyInto added in v1.5.5 

func ParseSeriesKeyInto(data []byte, dstTags models.Tags) ([]byte, models.Tags)

ParseSeriesKeyInto extracts the name and tags for data, parsing the tags into dstTags, which is then returened.

The returned dstTags may have a different length and capacity.

func ParseSeriesSegmentFilename added in v1.5.0 

func ParseSeriesSegmentFilename(filename string) (uint16, error)

ParseSeriesSegmentFilename returns the id represented by the hexidecimal filename.

func ReadAllSeriesIDIterator added in v1.5.0 

func ReadAllSeriesIDIterator(itr SeriesIDIterator) ([]uint64, error)

ReadAllSeriesIDIterator returns all ids from the iterator.

func ReadSeriesEntry added in v1.5.0 

func ReadSeriesEntry(data []byte) (flag uint8, id uint64, key []byte, sz int64)

func ReadSeriesKey added in v1.5.0 

func ReadSeriesKey(data []byte) (key, remainder []byte)

ReadSeriesKey returns the series key from the beginning of the buffer.

func ReadSeriesKeyFromSegments added in v1.5.0 

func ReadSeriesKeyFromSegments(a []*SeriesSegment, offset int64) []byte

ReadSeriesKeyFromSegments returns a series key from an offset within a set of segments.

func ReadSeriesKeyLen added in v1.5.0 

func ReadSeriesKeyLen(data []byte) (sz int, remainder []byte)

func ReadSeriesKeyMeasurement added in v1.5.0 

func ReadSeriesKeyMeasurement(data []byte) (name, remainder []byte)

func ReadSeriesKeyTag added in v1.5.0 

func ReadSeriesKeyTag(data []byte) (key, value, remainder []byte)

func ReadSeriesKeyTagN added in v1.5.0 

func ReadSeriesKeyTagN(data []byte) (n int, remainder []byte)

func RegisterEngine added in v0.9.3 

func RegisterEngine(name string, fn NewEngineFunc)

RegisterEngine registers a storage engine initializer by name.

func RegisterIndex added in v1.3.0 

func RegisterIndex(name string, fn NewIndexFunc)

RegisterIndex registers a storage index initializer by name.

func RegisteredEngines added in v0.9.5 

func RegisteredEngines() []string

RegisteredEngines returns the slice of currently registered engines.

func RegisteredIndexes added in v1.3.0 

func RegisteredIndexes() []string

RegisteredIndexes returns the slice of currently registered indexes.

func SeriesKeySize added in v1.5.0 

func SeriesKeySize(name []byte, tags models.Tags) int

SeriesKeySize returns the number of bytes required to encode a series key.

func SeriesKeysSize added in v1.5.0 

func SeriesKeysSize(names [][]byte, tagsSlice []models.Tags) int

SeriesKeysSize returns the number of bytes required to encode a list of name/tags.

func SeriesSegmentSize added in v1.5.0 

func SeriesSegmentSize(id uint16) uint32

SeriesSegmentSize returns the maximum size of the segment. The size goes up by powers of 2 starting from 4MB and reaching 256MB.

func SplitSeriesOffset added in v1.5.0 

func SplitSeriesOffset(offset int64) (segmentID uint16, pos uint32)

SplitSeriesOffset splits a offset into its 2-byte segmentID and 4-byte pos parts.

Types

type BooleanArray added in v1.7.0 

type BooleanArray = cursors.BooleanArray

func NewBooleanArrayLen added in v1.7.0 

func NewBooleanArrayLen(sz int) *BooleanArray

type BooleanArrayCursor added in v1.7.0 

type BooleanArrayCursor = cursors.BooleanArrayCursor

type CompactionPlannerCreator added in v1.6.0 

type CompactionPlannerCreator func(cfg Config) interface{}

type Config 

type Config struct {
	Dir    string `toml:"dir"`
	Engine string `toml:"-"`
	Index  string `toml:"index-version"`

	// General WAL configuration options
	WALDir string `toml:"wal-dir"`

	// WALFsyncDelay is the amount of time that a write will wait before fsyncing.  A duration
	// greater than 0 can be used to batch up multiple fsync calls.  This is useful for slower
	// disks or when WAL write contention is seen.  A value of 0 fsyncs every write to the WAL.
	WALFsyncDelay toml.Duration `toml:"wal-fsync-delay"`

	// Enables unicode validation on series keys on write.
	ValidateKeys bool `toml:"validate-keys"`

	// Query logging
	QueryLogEnabled bool `toml:"query-log-enabled"`

	// Compaction options for tsm1 (descriptions above with defaults)
	CacheMaxMemorySize             toml.Size     `toml:"cache-max-memory-size"`
	CacheSnapshotMemorySize        toml.Size     `toml:"cache-snapshot-memory-size"`
	CacheSnapshotWriteColdDuration toml.Duration `toml:"cache-snapshot-write-cold-duration"`
	CompactFullWriteColdDuration   toml.Duration `toml:"compact-full-write-cold-duration"`
	CompactThroughput              toml.Size     `toml:"compact-throughput"`
	CompactThroughputBurst         toml.Size     `toml:"compact-throughput-burst"`

	// MaxSeriesPerDatabase is the maximum number of series a node can hold per database.
	// When this limit is exceeded, writes return a 'max series per database exceeded' error.
	// A value of 0 disables the limit. This limit only applies when using the "inmem" index.
	MaxSeriesPerDatabase int `toml:"max-series-per-database"`

	// MaxValuesPerTag is the maximum number of tag values a single tag key can have within
	// a measurement.  When the limit is execeeded, writes return an error.
	// A value of 0 disables the limit.
	MaxValuesPerTag int `toml:"max-values-per-tag"`

	// MaxConcurrentCompactions is the maximum number of concurrent level and full compactions
	// that can be running at one time across all shards.  Compactions scheduled to run when the
	// limit is reached are blocked until a running compaction completes.  Snapshot compactions are
	// not affected by this limit.  A value of 0 limits compactions to runtime.GOMAXPROCS(0).
	MaxConcurrentCompactions int `toml:"max-concurrent-compactions"`

	// MaxIndexLogFileSize is the threshold, in bytes, when an index write-ahead log file will
	// compact into an index file. Lower sizes will cause log files to be compacted more quickly
	// and result in lower heap usage at the expense of write throughput. Higher sizes will
	// be compacted less frequently, store more series in-memory, and provide higher write throughput.
	MaxIndexLogFileSize toml.Size `toml:"max-index-log-file-size"`

	TraceLoggingEnabled bool `toml:"trace-logging-enabled"`

	// TSMWillNeed controls whether we hint to the kernel that we intend to
	// page in mmap'd sections of TSM files. This setting defaults to off, as it has
	// been found to be problematic in some cases. It may help users who have
	// slow disks.
	TSMWillNeed bool `toml:"tsm-use-madv-willneed"`
}

Config holds the configuration for the tsbd package.

func NewConfig 

func NewConfig() Config

NewConfig returns the default configuration for tsdb.

func (Config) Diagnostics added in v1.3.0 

func (c Config) Diagnostics() (*diagnostics.Diagnostics, error)

Diagnostics returns a diagnostics representation of a subset of the Config.

func (*Config) Validate added in v0.9.5 

func (c *Config) Validate() error

Validate validates the configuration hold by c.

type Cursor added in v0.9.3 

type Cursor = cursors.Cursor

type CursorIterator added in v1.6.0 

type CursorIterator = cursors.CursorIterator

type CursorIterators added in v1.6.0 

type CursorIterators = cursors.CursorIterators

func CreateCursorIterators added in v1.6.0 

func CreateCursorIterators(ctx context.Context, shards []*Shard) (CursorIterators, error)

type CursorRequest added in v1.4.0 

type CursorRequest = cursors.CursorRequest

type CursorStats added in v1.7.2 

type CursorStats = cursors.CursorStats

type Engine added in v0.9.3 

type Engine interface {
	Open() error
	Close() error
	SetEnabled(enabled bool)
	SetCompactionsEnabled(enabled bool)
	ScheduleFullCompaction() error

	WithLogger(*zap.Logger)

	LoadMetadataIndex(shardID uint64, index Index) error

	CreateSnapshot() (string, error)
	Backup(w io.Writer, basePath string, since time.Time) error
	Export(w io.Writer, basePath string, start time.Time, end time.Time) error
	Restore(r io.Reader, basePath string) error
	Import(r io.Reader, basePath string) error
	Digest() (io.ReadCloser, int64, error)

	CreateIterator(ctx context.Context, measurement string, opt query.IteratorOptions) (query.Iterator, error)
	CreateCursorIterator(ctx context.Context) (CursorIterator, error)
	IteratorCost(measurement string, opt query.IteratorOptions) (query.IteratorCost, error)
	WritePoints(points []models.Point) error

	CreateSeriesIfNotExists(key, name []byte, tags models.Tags) error
	CreateSeriesListIfNotExists(keys, names [][]byte, tags []models.Tags) error
	DeleteSeriesRange(itr SeriesIterator, min, max int64) error
	DeleteSeriesRangeWithPredicate(itr SeriesIterator, predicate func(name []byte, tags models.Tags) (int64, int64, bool)) error

	MeasurementsSketches() (estimator.Sketch, estimator.Sketch, error)
	SeriesSketches() (estimator.Sketch, estimator.Sketch, error)
	SeriesN() int64

	MeasurementExists(name []byte) (bool, error)

	MeasurementNamesByRegex(re *regexp.Regexp) ([][]byte, error)
	MeasurementFieldSet() *MeasurementFieldSet
	MeasurementFields(measurement []byte) *MeasurementFields
	ForEachMeasurementName(fn func(name []byte) error) error
	DeleteMeasurement(name []byte) error

	HasTagKey(name, key []byte) (bool, error)
	MeasurementTagKeysByExpr(name []byte, expr influxql.Expr) (map[string]struct{}, error)
	TagKeyCardinality(name, key []byte) int

	// Statistics will return statistics relevant to this engine.
	Statistics(tags map[string]string) []models.Statistic
	LastModified() time.Time
	DiskSize() int64
	IsIdle() bool
	Free() error

	io.WriterTo
}

Engine represents a swappable storage engine for the shard.

func NewEngine added in v0.9.3 

func NewEngine(id uint64, i Index, path string, walPath string, sfile *SeriesFile, options EngineOptions) (Engine, error)

NewEngine returns an instance of an engine based on its format. If the path does not exist then the DefaultFormat is used.

type EngineFormat added in v0.9.5 

type EngineFormat int

EngineFormat represents the format for an engine. 

const (
	// TSM1Format is the format used by the tsm1 engine.
	TSM1Format EngineFormat = 2
)

type EngineOptions added in v0.9.3 

type EngineOptions struct {
	EngineVersion string
	IndexVersion  string
	ShardID       uint64
	InmemIndex    interface{} // shared in-memory index

	// Limits the concurrent number of TSM files that can be loaded at once.
	OpenLimiter limiter.Fixed

	// CompactionDisabled specifies shards should not schedule compactions.
	// This option is intended for offline tooling.
	CompactionDisabled          bool
	CompactionPlannerCreator    CompactionPlannerCreator
	CompactionLimiter           limiter.Fixed
	CompactionThroughputLimiter limiter.Rate
	WALEnabled                  bool
	MonitorDisabled             bool

	// DatabaseFilter is a predicate controlling which databases may be opened.
	// If no function is set, all databases will be opened.
	DatabaseFilter func(database string) bool

	// RetentionPolicyFilter is a predicate controlling which combination of database and retention policy may be opened.
	// nil will allow all combinations to pass.
	RetentionPolicyFilter func(database, rp string) bool

	// ShardFilter is a predicate controlling which combination of database, retention policy and shard group may be opened.
	// nil will allow all combinations to pass.
	ShardFilter func(database, rp string, id uint64) bool

	Config         Config
	SeriesIDSets   SeriesIDSets
	FieldValidator FieldValidator

	OnNewEngine func(Engine)

	FileStoreObserver FileStoreObserver
}

EngineOptions represents the options used to initialize the engine.

func NewEngineOptions added in v0.9.3 

func NewEngineOptions() EngineOptions

NewEngineOptions constructs an EngineOptions object with safe default values. This should only be used in tests; production environments should read from a config file.

type Field added in v0.9.3 

type Field struct {
	ID   uint8             `json:"id,omitempty"`
	Name string            `json:"name,omitempty"`
	Type influxql.DataType `json:"type,omitempty"`
}

Field represents a series field.

type FieldCreate added in v0.9.3 

type FieldCreate struct {
	Measurement []byte
	Field       *Field
}

FieldCreate holds information for a field to create on a measurement.

type FieldValidator added in v1.6.0 

type FieldValidator interface {
	Validate(mf *MeasurementFields, point models.Point) error
}

FieldValidator should return a PartialWriteError if the point should not be written.

type FileStoreObserver added in v1.6.0 

type FileStoreObserver interface {
	// FileFinishing is called before a file is renamed to it's final name.
	FileFinishing(path string) error

	// FileUnlinking is called before a file is unlinked.
	FileUnlinking(path string) error
}

FileStoreObserver is passed notifications before the file store adds or deletes files. In this way, it can be sure to observe every file that is added or removed even in the presence of process death.

type FloatArray added in v1.7.0 

type FloatArray = cursors.FloatArray

func NewFloatArrayLen added in v1.7.0 

func NewFloatArrayLen(sz int) *FloatArray

type FloatArrayCursor added in v1.7.0 

type FloatArrayCursor = cursors.FloatArrayCursor

type Index added in v1.3.0 

type Index interface {
	Open() error
	Close() error
	WithLogger(*zap.Logger)

	Database() string
	MeasurementExists(name []byte) (bool, error)
	MeasurementNamesByRegex(re *regexp.Regexp) ([][]byte, error)
	DropMeasurement(name []byte) error
	ForEachMeasurementName(fn func(name []byte) error) error

	InitializeSeries(keys, names [][]byte, tags []models.Tags) error
	CreateSeriesIfNotExists(key, name []byte, tags models.Tags) error
	CreateSeriesListIfNotExists(keys, names [][]byte, tags []models.Tags) error
	DropSeries(seriesID uint64, key []byte, cascade bool) error
	DropMeasurementIfSeriesNotExist(name []byte) (bool, error)

	// Used to clean up series in inmem index that were dropped with a shard.
	DropSeriesGlobal(key []byte) error

	MeasurementsSketches() (estimator.Sketch, estimator.Sketch, error)
	SeriesN() int64
	SeriesSketches() (estimator.Sketch, estimator.Sketch, error)
	SeriesIDSet() *SeriesIDSet

	HasTagKey(name, key []byte) (bool, error)
	HasTagValue(name, key, value []byte) (bool, error)

	MeasurementTagKeysByExpr(name []byte, expr influxql.Expr) (map[string]struct{}, error)

	TagKeyCardinality(name, key []byte) int

	// InfluxQL system iterators
	MeasurementIterator() (MeasurementIterator, error)
	TagKeyIterator(name []byte) (TagKeyIterator, error)
	TagValueIterator(name, key []byte) (TagValueIterator, error)
	MeasurementSeriesIDIterator(name []byte) (SeriesIDIterator, error)
	TagKeySeriesIDIterator(name, key []byte) (SeriesIDIterator, error)
	TagValueSeriesIDIterator(name, key, value []byte) (SeriesIDIterator, error)

	// Sets a shared fieldset from the engine.
	FieldSet() *MeasurementFieldSet
	SetFieldSet(fs *MeasurementFieldSet)

	// Size of the index on disk, if applicable.
	DiskSizeBytes() int64

	// Bytes estimates the memory footprint of this Index, in bytes.
	Bytes() int

	// To be removed w/ tsi1.
	SetFieldName(measurement []byte, name string)

	Type() string
	// Returns a unique reference ID to the index instance.
	// For inmem, returns a reference to the backing Index, not ShardIndex.
	UniqueReferenceID() uintptr

	Rebuild()
}

func MustOpenIndex added in v1.3.0 

func MustOpenIndex(id uint64, database, path string, seriesIDSet *SeriesIDSet, sfile *SeriesFile, options EngineOptions) Index

func NewIndex added in v1.3.0 

func NewIndex(id uint64, database, path string, seriesIDSet *SeriesIDSet, sfile *SeriesFile, options EngineOptions) (Index, error)

NewIndex returns an instance of an index based on its format. If the path does not exist then the DefaultFormat is used.

type IndexFormat added in v1.3.0 

type IndexFormat int

IndexFormat represents the format for an index. 

const (
	// InMemFormat is the format used by the original in-memory shared index.
	InMemFormat IndexFormat = 1

	// TSI1Format is the format used by the tsi1 index.
	TSI1Format IndexFormat = 2
)

type IndexSet added in v1.5.0 

type IndexSet struct {
	Indexes    []Index     // The set of indexes comprising this IndexSet.
	SeriesFile *SeriesFile // The Series File associated with the db for this set.
	// contains filtered or unexported fields
}

IndexSet represents a list of indexes, all belonging to one database.

func (IndexSet) Database added in v1.5.0 

func (is IndexSet) Database() string

Database returns the database name of the first index.

func (IndexSet) DedupeInmemIndexes added in v1.5.0 

func (is IndexSet) DedupeInmemIndexes() IndexSet

DedupeInmemIndexes returns an index set which removes duplicate indexes. Useful because inmem indexes are shared by shards per database.

func (IndexSet) ForEachMeasurementTagKey added in v1.5.0 

func (is IndexSet) ForEachMeasurementTagKey(name []byte, fn func(key []byte) error) error

ForEachMeasurementTagKey iterates over all tag keys in a measurement and applies the provided function.

func (IndexSet) HasField added in v1.5.1 

func (is IndexSet) HasField(measurement []byte, field string) bool

HasField determines if any of the field sets on the set of indexes in the IndexSet have the provided field for the provided measurement.

func (IndexSet) HasInmemIndex added in v1.6.0 

func (is IndexSet) HasInmemIndex() bool

HasInmemIndex returns true if any in-memory index is in use.

func (IndexSet) HasTagKey added in v1.5.0 

func (is IndexSet) HasTagKey(name, key []byte) (bool, error)

HasTagKey returns true if the tag key exists in any index for the provided measurement.

func (IndexSet) HasTagValue added in v1.5.0 

func (is IndexSet) HasTagValue(name, key, value []byte) (bool, error)

HasTagValue returns true if the tag value exists in any index for the provided measurement and tag key.

func (IndexSet) MatchTagValueSeriesIDIterator added in v1.5.0 

func (is IndexSet) MatchTagValueSeriesIDIterator(name, key []byte, value *regexp.Regexp, matches bool) (SeriesIDIterator, error)

MatchTagValueSeriesIDIterator returns a series iterator for tags which match value. If matches is false, returns iterators which do not match value.

func (IndexSet) MeasurementIterator added in v1.5.0 

func (is IndexSet) MeasurementIterator() (MeasurementIterator, error)

MeasurementIterator returns an iterator over all measurements in the index.

func (IndexSet) MeasurementNamesByExpr added in v1.5.0 

func (is IndexSet) MeasurementNamesByExpr(auth query.Authorizer, expr influxql.Expr) ([][]byte, error)

MeasurementNamesByExpr returns a slice of measurement names matching the provided condition. If no condition is provided then all names are returned.

func (IndexSet) MeasurementSeriesByExprIterator added in v1.5.0 

func (is IndexSet) MeasurementSeriesByExprIterator(name []byte, expr influxql.Expr) (SeriesIDIterator, error)

MeasurementSeriesByExprIterator returns a series iterator for a measurement that is filtered by expr. If expr only contains time expressions then this call is equivalent to MeasurementSeriesIDIterator().

func (IndexSet) MeasurementSeriesIDIterator added in v1.5.0 

func (is IndexSet) MeasurementSeriesIDIterator(name []byte) (SeriesIDIterator, error)

MeasurementSeriesIDIterator returns an iterator over all non-tombstoned series for the provided measurement.

func (IndexSet) MeasurementSeriesKeysByExpr added in v1.5.0 

func (is IndexSet) MeasurementSeriesKeysByExpr(name []byte, expr influxql.Expr) ([][]byte, error)

MeasurementSeriesKeysByExpr returns a list of series keys matching expr.

func (IndexSet) MeasurementTagKeyValuesByExpr added in v1.5.0 

func (is IndexSet) MeasurementTagKeyValuesByExpr(auth query.Authorizer, name []byte, keys []string, expr influxql.Expr, keysSorted bool) ([][]string, error)

MeasurementTagKeyValuesByExpr returns a set of tag values filtered by an expression.

func (IndexSet) MeasurementTagKeysByExpr added in v1.5.0 

func (is IndexSet) MeasurementTagKeysByExpr(name []byte, expr influxql.Expr) (map[string]struct{}, error)

MeasurementTagKeysByExpr extracts the tag keys wanted by the expression.

func (IndexSet) TagKeyHasAuthorizedSeries added in v1.5.0 

func (is IndexSet) TagKeyHasAuthorizedSeries(auth query.Authorizer, name, tagKey []byte) (bool, error)

TagKeyHasAuthorizedSeries determines if there exists an authorized series for the provided measurement name and tag key.

func (IndexSet) TagKeyIterator added in v1.5.0 

func (is IndexSet) TagKeyIterator(name []byte) (TagKeyIterator, error)

TagKeyIterator returns a key iterator for a measurement.

func (IndexSet) TagKeySeriesIDIterator added in v1.5.0 

func (is IndexSet) TagKeySeriesIDIterator(name, key []byte) (SeriesIDIterator, error)

TagKeySeriesIDIterator returns a series iterator for all values across a single key.

func (IndexSet) TagSets added in v1.5.0 

func (is IndexSet) TagSets(sfile *SeriesFile, name []byte, opt query.IteratorOptions) ([]*query.TagSet, error)

TagSets returns an ordered list of tag sets for a measurement by dimension and filtered by an optional conditional expression.

func (IndexSet) TagValueIterator added in v1.5.0 

func (is IndexSet) TagValueIterator(name, key []byte) (TagValueIterator, error)

TagValueIterator returns a value iterator for a tag key.

func (IndexSet) TagValueSeriesIDIterator added in v1.5.0 

func (is IndexSet) TagValueSeriesIDIterator(name, key, value []byte) (SeriesIDIterator, error)

TagValueSeriesIDIterator returns a series iterator for a single tag value.

func (IndexSet) TagValuesByKeyAndExpr added in v1.5.0 

func (is IndexSet) TagValuesByKeyAndExpr(auth query.Authorizer, name []byte, keys []string, expr influxql.Expr, fieldset *MeasurementFieldSet) ([]map[string]struct{}, error)

TagValuesByKeyAndExpr retrieves tag values for the provided tag keys.

TagValuesByKeyAndExpr returns sets of values for each key, indexable by the position of the tag key in the keys argument.

N.B tagValuesByKeyAndExpr relies on keys being sorted in ascending lexicographic order.

type IntegerArray added in v1.7.0 

type IntegerArray = cursors.IntegerArray

func NewIntegerArrayLen added in v1.7.0 

func NewIntegerArrayLen(sz int) *IntegerArray

type IntegerArrayCursor added in v1.7.0 

type IntegerArrayCursor = cursors.IntegerArrayCursor

type KeyValue added in v1.0.0 

type KeyValue struct {
	Key, Value string
}

KeyValue holds a string key and a string value.

type KeyValues added in v1.0.0 

type KeyValues []KeyValue

KeyValues is a sortable slice of KeyValue.

func (KeyValues) Len added in v1.0.0 

func (a KeyValues) Len() int

Len implements sort.Interface.

func (KeyValues) Less added in v1.0.0 

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

Less implements sort.Interface. Keys are compared before values.

func (KeyValues) Swap added in v1.0.0 

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

Swap implements sort.Interface.

type LimitError added in v1.3.0 

type LimitError struct {
	Reason string
}

LimitError represents an error caused by a configurable limit.

func (*LimitError) Error added in v1.3.0 

func (e *LimitError) Error() string

type MeasurementFieldSet added in v1.3.0 

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

MeasurementFieldSet represents a collection of fields by measurement. This safe for concurrent use.

func NewMeasurementFieldSet added in v1.3.0 

func NewMeasurementFieldSet(path string) (*MeasurementFieldSet, error)

NewMeasurementFieldSet returns a new instance of MeasurementFieldSet.

func (*MeasurementFieldSet) Bytes added in v1.6.0 

func (fs *MeasurementFieldSet) Bytes() int

Bytes estimates the memory footprint of this MeasurementFieldSet, in bytes.

func (*MeasurementFieldSet) CreateFieldsIfNotExists added in v1.3.0 

func (fs *MeasurementFieldSet) CreateFieldsIfNotExists(name []byte) *MeasurementFields

CreateFieldsIfNotExists returns fields for a measurement by name.

func (*MeasurementFieldSet) Delete added in v1.3.0 

func (fs *MeasurementFieldSet) Delete(name string)

Delete removes a field set for a measurement.

func (*MeasurementFieldSet) DeleteWithLock added in v1.3.0 

func (fs *MeasurementFieldSet) DeleteWithLock(name string, fn func() error) error

DeleteWithLock executes fn and removes a field set from a measurement under lock.

func (*MeasurementFieldSet) Fields added in v1.3.0 

func (fs *MeasurementFieldSet) Fields(name []byte) *MeasurementFields

Fields returns fields for a measurement by name.

func (*MeasurementFieldSet) FieldsByString added in v1.5.1 

func (fs *MeasurementFieldSet) FieldsByString(name string) *MeasurementFields

FieldsByString returns fields for a measurment by name.

func (*MeasurementFieldSet) IsEmpty added in v1.5.0 

func (fs *MeasurementFieldSet) IsEmpty() bool

func (*MeasurementFieldSet) Save added in v1.5.0 

func (fs *MeasurementFieldSet) Save() error

type MeasurementFields added in v0.9.3 

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

MeasurementFields holds the fields of a measurement and their codec.

func NewMeasurementFields added in v0.12.1 

func NewMeasurementFields() *MeasurementFields

NewMeasurementFields returns an initialised *MeasurementFields value.

func (*MeasurementFields) CreateFieldIfNotExists added in v0.9.3 

func (m *MeasurementFields) CreateFieldIfNotExists(name []byte, typ influxql.DataType) error

CreateFieldIfNotExists creates a new field with an autoincrementing ID. Returns an error if 255 fields have already been created on the measurement or the fields already exists with a different type.

func (*MeasurementFields) Field added in v0.12.1 

func (m *MeasurementFields) Field(name string) *Field

Field returns the field for name, or nil if there is no field for name.

func (*MeasurementFields) FieldBytes added in v1.1.0 

func (m *MeasurementFields) FieldBytes(name []byte) *Field

FieldBytes returns the field for name, or nil if there is no field for name. FieldBytes should be preferred to Field when the caller has a []byte, because it avoids a string allocation, which can't be avoided if the caller converts the []byte to a string and calls Field.

func (*MeasurementFields) FieldKeys added in v1.4.0 

func (m *MeasurementFields) FieldKeys() []string

func (*MeasurementFields) FieldN added in v1.3.0 

func (m *MeasurementFields) FieldN() int

func (*MeasurementFields) FieldSet added in v1.0.0 

func (m *MeasurementFields) FieldSet() map[string]influxql.DataType

FieldSet returns the set of fields and their types for the measurement.

func (*MeasurementFields) ForEachField added in v1.5.0 

func (m *MeasurementFields) ForEachField(fn func(name string, typ influxql.DataType) bool)

func (*MeasurementFields) HasField added in v1.3.0 

func (m *MeasurementFields) HasField(name string) bool

type MeasurementIterator added in v0.11.0 

type MeasurementIterator interface {
	Close() error
	Next() ([]byte, error)
}

MeasurementIterator represents a iterator over a list of measurements.

func MergeMeasurementIterators added in v1.5.0 

func MergeMeasurementIterators(itrs ...MeasurementIterator) MeasurementIterator

MergeMeasurementIterators returns an iterator that merges a set of iterators. Iterators that are first in the list take precendence and a deletion by those early iterators will invalidate elements by later iterators.

type MeasurementIterators added in v1.5.0 

type MeasurementIterators []MeasurementIterator

func (MeasurementIterators) Close added in v1.5.0 

func (a MeasurementIterators) Close() (err error)

type NewEngineFunc added in v0.9.3 

type NewEngineFunc func(id uint64, i Index, path string, walPath string, sfile *SeriesFile, options EngineOptions) Engine

NewEngineFunc creates a new engine.

type NewIndexFunc added in v1.3.0 

type NewIndexFunc func(id uint64, database, path string, seriesIDSet *SeriesIDSet, sfile *SeriesFile, options EngineOptions) Index

NewIndexFunc creates a new index.

type PartialWriteError added in v1.1.0 

type PartialWriteError struct {
	Reason  string
	Dropped int

	// A sorted slice of series keys that were dropped.
	DroppedKeys [][]byte
}

PartialWriteError indicates a write request could only write a portion of the requested values.

func (PartialWriteError) Error added in v1.1.0 

func (e PartialWriteError) Error() string

type PointBatcher 

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

PointBatcher accepts Points and will emit a batch of those points when either a) the batch reaches a certain size, or b) a certain time passes.

func NewPointBatcher 

func NewPointBatcher(sz int, bp int, d time.Duration) *PointBatcher

NewPointBatcher returns a new PointBatcher. sz is the batching size, bp is the maximum number of batches that may be pending. d is the time after which a batch will be emitted after the first point is received for the batch, regardless of its size.

func (*PointBatcher) Flush 

func (b *PointBatcher) Flush()

Flush instructs the batcher to emit any pending points in a batch, regardless of batch size. If there are no pending points, no batch is emitted.

func (*PointBatcher) In 

func (b *PointBatcher) In() chan<- models.Point

In returns the channel to which points should be written.

func (*PointBatcher) Out 

func (b *PointBatcher) Out() <-chan []models.Point

Out returns the channel from which batches should be read.

func (*PointBatcher) Start 

func (b *PointBatcher) Start()

Start starts the batching process. Returns the in and out channels for points and point-batches respectively.

func (*PointBatcher) Stats 

func (b *PointBatcher) Stats() *PointBatcherStats

Stats returns a PointBatcherStats object for the PointBatcher. While the each statistic should be closely correlated with each other statistic, it is not guaranteed.

func (*PointBatcher) Stop 

func (b *PointBatcher) Stop()

Stop stops the batching process. Stop waits for the batching routine to stop before returning.

type PointBatcherStats 

type PointBatcherStats struct {
	BatchTotal   uint64 // Total count of batches transmitted.
	PointTotal   uint64 // Total count of points processed.
	SizeTotal    uint64 // Number of batches that reached size threshold.
	TimeoutTotal uint64 // Number of timeouts that occurred.
}

PointBatcherStats are the statistics each batcher tracks.

type SeriesCursor added in v1.6.0 

type SeriesCursor interface {
	Close() error
	Next() (*SeriesCursorRow, error)
}

type SeriesCursorRequest added in v1.6.0 

type SeriesCursorRequest struct {
	Measurements MeasurementIterator
}

type SeriesCursorRow added in v1.6.0 

type SeriesCursorRow struct {
	Name []byte
	Tags models.Tags
}

func (*SeriesCursorRow) Compare added in v1.6.0 

func (r *SeriesCursorRow) Compare(other *SeriesCursorRow) int

type SeriesElem added in v1.5.0 

type SeriesElem interface {
	Name() []byte
	Tags() models.Tags
	Deleted() bool

	// InfluxQL expression associated with series during filtering.
	Expr() influxql.Expr
}

SeriesElem represents a generic series element.

type SeriesFile added in v1.5.0 

type SeriesFile struct {
	Logger *zap.Logger
	// contains filtered or unexported fields
}

SeriesFile represents the section of the index that holds series data.

func NewSeriesFile added in v1.5.0 

func NewSeriesFile(path string) *SeriesFile

NewSeriesFile returns a new instance of SeriesFile.

func (*SeriesFile) Close added in v1.5.0 

func (f *SeriesFile) Close() (err error)

Close unmaps the data file.

func (*SeriesFile) CreateSeriesListIfNotExists added in v1.5.0 

func (f *SeriesFile) CreateSeriesListIfNotExists(names [][]byte, tagsSlice []models.Tags) ([]uint64, error)

CreateSeriesListIfNotExists creates a list of series in bulk if they don't exist. The returned ids slice returns IDs for every name+tags, creating new series IDs as needed.

func (*SeriesFile) DeleteSeriesID added in v1.5.0 

func (f *SeriesFile) DeleteSeriesID(id uint64) error

DeleteSeriesID flags a series as permanently deleted. If the series is reintroduced later then it must create a new id.

func (*SeriesFile) DisableCompactions added in v1.5.0 

func (f *SeriesFile) DisableCompactions()

DisableCompactions prevents new compactions from running.

func (*SeriesFile) EnableCompactions added in v1.5.0 

func (f *SeriesFile) EnableCompactions()

EnableCompactions allows compactions to run.

func (*SeriesFile) HasSeries added in v1.5.0 

func (f *SeriesFile) HasSeries(name []byte, tags models.Tags, buf []byte) bool

HasSeries return true if the series exists.

func (*SeriesFile) IsDeleted added in v1.5.0 

func (f *SeriesFile) IsDeleted(id uint64) bool

IsDeleted returns true if the ID has been deleted before.

func (*SeriesFile) Open added in v1.5.0 

func (f *SeriesFile) Open() error

Open memory maps the data file at the file's path.

func (*SeriesFile) Partitions added in v1.5.0 

func (f *SeriesFile) Partitions() []*SeriesPartition

Partitions returns all partitions.

func (*SeriesFile) Path added in v1.5.0 

func (f *SeriesFile) Path() string

Path returns the path to the file.

func (*SeriesFile) Retain added in v1.5.0 

func (f *SeriesFile) Retain() func()

Retain adds a reference count to the file. It returns a release func.

func (*SeriesFile) Series added in v1.5.0 

func (f *SeriesFile) Series(id uint64) ([]byte, models.Tags)

Series returns the parsed series name and tags for an offset.

func (*SeriesFile) SeriesCount added in v1.5.0 

func (f *SeriesFile) SeriesCount() uint64

SeriesCount returns the number of series.

func (*SeriesFile) SeriesID added in v1.5.0 

func (f *SeriesFile) SeriesID(name []byte, tags models.Tags, buf []byte) uint64

SeriesID return the series id for the series.

func (*SeriesFile) SeriesIDIterator added in v1.5.0 

func (f *SeriesFile) SeriesIDIterator() SeriesIDIterator

SeriesIterator returns an iterator over all the series.

func (*SeriesFile) SeriesIDPartition added in v1.5.0 

func (f *SeriesFile) SeriesIDPartition(id uint64) *SeriesPartition

func (*SeriesFile) SeriesIDPartitionID added in v1.5.0 

func (f *SeriesFile) SeriesIDPartitionID(id uint64) int

func (*SeriesFile) SeriesKey added in v1.5.0 

func (f *SeriesFile) SeriesKey(id uint64) []byte

SeriesKey returns the series key for a given id.

func (*SeriesFile) SeriesKeyPartition added in v1.5.0 

func (f *SeriesFile) SeriesKeyPartition(key []byte) *SeriesPartition

func (*SeriesFile) SeriesKeyPartitionID added in v1.5.0 

func (f *SeriesFile) SeriesKeyPartitionID(key []byte) int

func (*SeriesFile) SeriesKeys added in v1.5.0 

func (f *SeriesFile) SeriesKeys(ids []uint64) [][]byte

SeriesKeys returns a list of series keys from a list of ids.

func (*SeriesFile) SeriesKeysPartitionIDs added in v1.5.0 

func (f *SeriesFile) SeriesKeysPartitionIDs(keys [][]byte) []int

func (*SeriesFile) SeriesPartitionPath added in v1.5.0 

func (f *SeriesFile) SeriesPartitionPath(i int) string

SeriesPartitionPath returns the path to a given partition.

func (*SeriesFile) Wait added in v1.5.0 

func (f *SeriesFile) Wait()

Wait waits for all Retains to be released.

type SeriesIDElem added in v1.5.0 

type SeriesIDElem struct {
	SeriesID uint64
	Expr     influxql.Expr
}

SeriesIDElem represents a single series and optional expression.

type SeriesIDElems added in v1.5.0 

type SeriesIDElems []SeriesIDElem

SeriesIDElems represents a list of series id elements.

func (SeriesIDElems) Len added in v1.5.0 

func (a SeriesIDElems) Len() int

func (SeriesIDElems) Less added in v1.5.0 

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

func (SeriesIDElems) Swap added in v1.5.0 

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

type SeriesIDIterator added in v1.5.0 

type SeriesIDIterator interface {
	Next() (SeriesIDElem, error)
	Close() error
}

SeriesIDIterator represents a iterator over a list of series ids.

func DifferenceSeriesIDIterators added in v1.5.0 

func DifferenceSeriesIDIterators(itr0, itr1 SeriesIDIterator) SeriesIDIterator

DifferenceSeriesIDIterators returns an iterator that only returns series which occur the first iterator but not the second iterator.

func FilterUndeletedSeriesIDIterator added in v1.5.0 

func FilterUndeletedSeriesIDIterator(sfile *SeriesFile, itr SeriesIDIterator) SeriesIDIterator

FilterUndeletedSeriesIDIterator returns an iterator which filters all deleted series.

func IntersectSeriesIDIterators added in v1.5.0 

func IntersectSeriesIDIterators(itr0, itr1 SeriesIDIterator) SeriesIDIterator

IntersectSeriesIDIterators returns an iterator that only returns series which occur in both iterators. If both series have associated expressions then they are combined together.

func MergeSeriesIDIterators added in v1.5.0 

func MergeSeriesIDIterators(itrs ...SeriesIDIterator) SeriesIDIterator

MergeSeriesIDIterators returns an iterator that merges a set of iterators. Iterators that are first in the list take precedence and a deletion by those early iterators will invalidate elements by later iterators.

func UnionSeriesIDIterators added in v1.5.0 

func UnionSeriesIDIterators(itr0, itr1 SeriesIDIterator) SeriesIDIterator

UnionSeriesIDIterators returns an iterator that returns series from both both iterators. If both series have associated expressions then they are combined together.

type SeriesIDIterators added in v1.5.0 

type SeriesIDIterators []SeriesIDIterator

func (SeriesIDIterators) Close added in v1.5.0 

func (a SeriesIDIterators) Close() (err error)

type SeriesIDSet added in v1.5.0 

type SeriesIDSet struct {
	sync.RWMutex
	// contains filtered or unexported fields
}

SeriesIDSet represents a lockable bitmap of series ids.

func NewSeriesIDSet added in v1.5.0 

func NewSeriesIDSet(a ...uint64) *SeriesIDSet

NewSeriesIDSet returns a new instance of SeriesIDSet.

func (*SeriesIDSet) Add added in v1.5.0 

func (s *SeriesIDSet) Add(id uint64)

Add adds the series id to the set.

func (*SeriesIDSet) AddMany added in v1.7.0 

func (s *SeriesIDSet) AddMany(ids ...uint64)

AddMany adds multiple ids to the SeriesIDSet. AddMany takes a lock, so may not be optimal to call many times with few ids.

func (*SeriesIDSet) AddNoLock added in v1.5.0 

func (s *SeriesIDSet) AddNoLock(id uint64)

AddNoLock adds the series id to the set. Add is not safe for use from multiple goroutines. Callers must manage synchronization.

func (*SeriesIDSet) And added in v1.6.1 

func (s *SeriesIDSet) And(other *SeriesIDSet) *SeriesIDSet

And returns a new SeriesIDSet containing elements that were present in s and other.

func (*SeriesIDSet) AndNot added in v1.5.0 

func (s *SeriesIDSet) AndNot(other *SeriesIDSet) *SeriesIDSet

AndNot returns a new SeriesIDSet containing elements that were present in s, but not present in other.

func (*SeriesIDSet) Bytes added in v1.6.0 

func (s *SeriesIDSet) Bytes() int

Bytes estimates the memory footprint of this SeriesIDSet, in bytes.

func (*SeriesIDSet) Cardinality added in v1.5.0 

func (s *SeriesIDSet) Cardinality() uint64

Cardinality returns the cardinality of the SeriesIDSet.

func (*SeriesIDSet) Clear added in v1.7.0 

func (s *SeriesIDSet) Clear()

Clear clears the underlying bitmap for re-use. Clear is safe for use by multiple goroutines.

func (*SeriesIDSet) ClearNoLock added in v1.7.0 

func (s *SeriesIDSet) ClearNoLock()

ClearNoLock clears the underlying bitmap for re-use without taking a lock.

func (*SeriesIDSet) Clone added in v1.6.1 

func (s *SeriesIDSet) Clone() *SeriesIDSet

Clone returns a new SeriesIDSet with a deep copy of the underlying bitmap.

func (*SeriesIDSet) CloneNoLock added in v1.6.1 

func (s *SeriesIDSet) CloneNoLock() *SeriesIDSet

CloneNoLock calls Clone without taking a lock.

func (*SeriesIDSet) Contains added in v1.5.0 

func (s *SeriesIDSet) Contains(id uint64) bool

Contains returns true if the id exists in the set.

func (*SeriesIDSet) ContainsNoLock added in v1.5.0 

func (s *SeriesIDSet) ContainsNoLock(id uint64) bool

ContainsNoLock returns true if the id exists in the set. ContainsNoLock is not safe for use from multiple goroutines. The caller must manage synchronization.

func (*SeriesIDSet) Diff added in v1.5.0 

func (s *SeriesIDSet) Diff(other *SeriesIDSet)

Diff removes from s any elements also present in other.

func (*SeriesIDSet) Equals added in v1.5.0 

func (s *SeriesIDSet) Equals(other *SeriesIDSet) bool

Equals returns true if other and s are the same set of ids.

func (*SeriesIDSet) ForEach added in v1.5.0 

func (s *SeriesIDSet) ForEach(f func(id uint64))

ForEach calls f for each id in the set. The function is applied to the IDs in ascending order.

func (*SeriesIDSet) ForEachNoLock added in v1.6.1 

func (s *SeriesIDSet) ForEachNoLock(f func(id uint64))

ForEachNoLock calls f for each id in the set without taking a lock.

func (*SeriesIDSet) Iterator added in v1.6.1 

func (s *SeriesIDSet) Iterator() SeriesIDSetIterable

Iterator returns an iterator to the underlying bitmap. This iterator is not protected by a lock.

func (*SeriesIDSet) Merge added in v1.5.0 

func (s *SeriesIDSet) Merge(others ...*SeriesIDSet)

Merge merged the contents of others into s. The caller does not need to provide s as an argument, and the contents of s will always be present in s after Merge returns.

func (*SeriesIDSet) MergeInPlace added in v1.7.0 

func (s *SeriesIDSet) MergeInPlace(other *SeriesIDSet)

MergeInPlace merges other into s, modifying s in the process.

func (*SeriesIDSet) Remove added in v1.5.0 

func (s *SeriesIDSet) Remove(id uint64)

Remove removes the id from the set.

func (*SeriesIDSet) RemoveNoLock added in v1.5.0 

func (s *SeriesIDSet) RemoveNoLock(id uint64)

RemoveNoLock removes the id from the set. RemoveNoLock is not safe for use from multiple goroutines. The caller must manage synchronization.

func (*SeriesIDSet) SetCOW added in v1.7.0 

func (s *SeriesIDSet) SetCOW(b bool)

SetCOW sets the copy-on-write status of the underlying bitmap. When SetCOW(true) is called, modifications to the bitmap will result in copies of the relevant data structures being made, preventing consumers of clones of the bitmap from seeing those modifications.

func (*SeriesIDSet) Slice added in v1.7.0 

func (s *SeriesIDSet) Slice() []uint64

Slice returns a slice of series ids.

func (*SeriesIDSet) String added in v1.5.0 

func (s *SeriesIDSet) String() string

func (*SeriesIDSet) UnmarshalBinary added in v1.5.0 

func (s *SeriesIDSet) UnmarshalBinary(data []byte) error

UnmarshalBinary unmarshals data into the set.

func (*SeriesIDSet) UnmarshalBinaryUnsafe added in v1.7.0 

func (s *SeriesIDSet) UnmarshalBinaryUnsafe(data []byte) error

UnmarshalBinaryUnsafe unmarshals data into the set. References to the underlying data are used so data should not be reused by caller.

func (*SeriesIDSet) WriteTo added in v1.5.0 

func (s *SeriesIDSet) WriteTo(w io.Writer) (int64, error)

WriteTo writes the set to w.

type SeriesIDSetIterable added in v1.6.1 

type SeriesIDSetIterable interface {
	HasNext() bool
	Next() uint32
}

type SeriesIDSetIterator added in v1.6.1 

type SeriesIDSetIterator interface {
	SeriesIDIterator
	SeriesIDSet() *SeriesIDSet
}

SeriesIDSetIterator represents an iterator that can produce a SeriesIDSet.

func NewSeriesIDSetIterator added in v1.6.1 

func NewSeriesIDSetIterator(ss *SeriesIDSet) SeriesIDSetIterator

func NewSeriesIDSetIterators added in v1.6.1 

func NewSeriesIDSetIterators(itrs []SeriesIDIterator) []SeriesIDSetIterator

NewSeriesIDSetIterators returns a slice of SeriesIDSetIterator if all itrs can be type casted. Otherwise returns nil.

type SeriesIDSets added in v1.5.0 

type SeriesIDSets interface {
	ForEach(f func(ids *SeriesIDSet)) error
}

SeriesIDSets provides access to the total set of series IDs

type SeriesIDSliceIterator added in v1.5.0 

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

SeriesIDSliceIterator iterates over a slice of series ids.

func NewSeriesIDSliceIterator added in v1.5.0 

func NewSeriesIDSliceIterator(ids []uint64) *SeriesIDSliceIterator

NewSeriesIDSliceIterator returns a SeriesIDIterator that iterates over a slice.

func (*SeriesIDSliceIterator) Close added in v1.5.0 

func (itr *SeriesIDSliceIterator) Close() error

func (*SeriesIDSliceIterator) Next added in v1.5.0 

func (itr *SeriesIDSliceIterator) Next() (SeriesIDElem, error)

Next returns the next series id in the slice.

func (*SeriesIDSliceIterator) SeriesIDSet added in v1.6.1 

func (itr *SeriesIDSliceIterator) SeriesIDSet() *SeriesIDSet

SeriesIDSet returns a set of all remaining ids.

type SeriesIndex added in v1.5.0 

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

SeriesIndex represents an index of key-to-id & id-to-offset mappings.

func NewSeriesIndex added in v1.5.0 

func NewSeriesIndex(path string) *SeriesIndex

func (*SeriesIndex) Clone added in v1.5.0 

func (idx *SeriesIndex) Clone() *SeriesIndex

Clone returns a copy of idx for use during compaction. In-memory maps are not cloned.

func (*SeriesIndex) Close added in v1.5.0 

func (idx *SeriesIndex) Close() (err error)

Close unmaps the index file.

func (*SeriesIndex) Count added in v1.5.0 

func (idx *SeriesIndex) Count() uint64

Count returns the number of series in the index.

func (*SeriesIndex) Delete added in v1.5.0 

func (idx *SeriesIndex) Delete(id uint64)

Delete marks the series id as deleted.

func (*SeriesIndex) FindIDByNameTags added in v1.5.0 

func (idx *SeriesIndex) FindIDByNameTags(segments []*SeriesSegment, name []byte, tags models.Tags, buf []byte) uint64

func (*SeriesIndex) FindIDBySeriesKey added in v1.5.0 

func (idx *SeriesIndex) FindIDBySeriesKey(segments []*SeriesSegment, key []byte) uint64

func (*SeriesIndex) FindIDListByNameTags added in v1.5.0 

func (idx *SeriesIndex) FindIDListByNameTags(segments []*SeriesSegment, names [][]byte, tagsSlice []models.Tags, buf []byte) (ids []uint64, ok bool)

func (*SeriesIndex) FindOffsetByID added in v1.5.0 

func (idx *SeriesIndex) FindOffsetByID(id uint64) int64

func (*SeriesIndex) InMemCount added in v1.5.0 

func (idx *SeriesIndex) InMemCount() uint64

InMemCount returns the number of series in the in-memory index.

func (*SeriesIndex) Insert added in v1.5.0 

func (idx *SeriesIndex) Insert(key []byte, id uint64, offset int64)

func (*SeriesIndex) IsDeleted added in v1.5.0 

func (idx *SeriesIndex) IsDeleted(id uint64) bool

IsDeleted returns true if series id has been deleted.

func (*SeriesIndex) OnDiskCount added in v1.5.0 

func (idx *SeriesIndex) OnDiskCount() uint64

OnDiskCount returns the number of series in the on-disk index.

func (*SeriesIndex) Open added in v1.5.0 

func (idx *SeriesIndex) Open() (err error)

Open memory-maps the index file.

func (*SeriesIndex) Recover added in v1.5.0 

func (idx *SeriesIndex) Recover(segments []*SeriesSegment) error

Recover rebuilds the in-memory index for all new entries.

type SeriesIndexHeader added in v1.5.0 

type SeriesIndexHeader struct {
	Version uint8

	MaxSeriesID uint64
	MaxOffset   int64

	Count    uint64
	Capacity int64

	KeyIDMap struct {
		Offset int64
		Size   int64
	}

	IDOffsetMap struct {
		Offset int64
		Size   int64
	}
}

SeriesIndexHeader represents the header of a series index.

func NewSeriesIndexHeader added in v1.5.0 

func NewSeriesIndexHeader() SeriesIndexHeader

NewSeriesIndexHeader returns a new instance of SeriesIndexHeader.

func ReadSeriesIndexHeader added in v1.5.0 

func ReadSeriesIndexHeader(data []byte) (hdr SeriesIndexHeader, err error)

ReadSeriesIndexHeader returns the header from data.

func (*SeriesIndexHeader) WriteTo added in v1.5.0 

func (hdr *SeriesIndexHeader) WriteTo(w io.Writer) (n int64, err error)

WriteTo writes the header to w.

type SeriesIterator added in v1.5.0 

type SeriesIterator interface {
	Close() error
	Next() (SeriesElem, error)
}

SeriesIterator represents a iterator over a list of series.

func NewSeriesIteratorAdapter added in v1.5.0 

func NewSeriesIteratorAdapter(sfile *SeriesFile, itr SeriesIDIterator) SeriesIterator

NewSeriesIteratorAdapter returns an adapter for converting series ids to series.

type SeriesPartition added in v1.5.0 

type SeriesPartition struct {
	CompactThreshold int

	Logger *zap.Logger
	// contains filtered or unexported fields
}

SeriesPartition represents a subset of series file data.

func NewSeriesPartition added in v1.5.0 

func NewSeriesPartition(id int, path string) *SeriesPartition

NewSeriesPartition returns a new instance of SeriesPartition.

func (*SeriesPartition) AppendSeriesIDs added in v1.5.0 

func (p *SeriesPartition) AppendSeriesIDs(a []uint64) []uint64

AppendSeriesIDs returns a list of all series ids.

func (*SeriesPartition) Close added in v1.5.0 

func (p *SeriesPartition) Close() (err error)

Close unmaps the data files.

func (*SeriesPartition) Compacting added in v1.6.0 

func (p *SeriesPartition) Compacting() bool

Compacting returns if the SeriesPartition is currently compacting.

func (*SeriesPartition) CreateSeriesListIfNotExists added in v1.5.0 

func (p *SeriesPartition) CreateSeriesListIfNotExists(keys [][]byte, keyPartitionIDs []int, ids []uint64) error

CreateSeriesListIfNotExists creates a list of series in bulk if they don't exist. The ids parameter is modified to contain series IDs for all keys belonging to this partition.

func (*SeriesPartition) DeleteSeriesID added in v1.5.0 

func (p *SeriesPartition) DeleteSeriesID(id uint64) error

DeleteSeriesID flags a series as permanently deleted. If the series is reintroduced later then it must create a new id.

func (*SeriesPartition) DisableCompactions added in v1.5.0 

func (p *SeriesPartition) DisableCompactions()

func (*SeriesPartition) EnableCompactions added in v1.5.0 

func (p *SeriesPartition) EnableCompactions()

func (*SeriesPartition) FindIDBySeriesKey added in v1.5.0 

func (p *SeriesPartition) FindIDBySeriesKey(key []byte) uint64

FindIDBySeriesKey return the series id for the series key.

func (*SeriesPartition) ID added in v1.5.0 

func (p *SeriesPartition) ID() int

ID returns the partition id.

func (*SeriesPartition) IndexPath added in v1.5.0 

func (p *SeriesPartition) IndexPath() string

Path returns the path to the series index.

func (*SeriesPartition) IsDeleted added in v1.5.0 

func (p *SeriesPartition) IsDeleted(id uint64) bool

IsDeleted returns true if the ID has been deleted before.

func (*SeriesPartition) Open added in v1.5.0 

func (p *SeriesPartition) Open() error

Open memory maps the data file at the partition's path.

func (*SeriesPartition) Path added in v1.5.0 

func (p *SeriesPartition) Path() string

Path returns the path to the partition.

func (*SeriesPartition) Series added in v1.5.0 

func (p *SeriesPartition) Series(id uint64) ([]byte, models.Tags)

Series returns the parsed series name and tags for an offset.

func (*SeriesPartition) SeriesCount added in v1.5.0 

func (p *SeriesPartition) SeriesCount() uint64

SeriesCount returns the number of series.

func (*SeriesPartition) SeriesKey added in v1.5.0 

func (p *SeriesPartition) SeriesKey(id uint64) []byte

SeriesKey returns the series key for a given id.

type SeriesPartitionCompactor added in v1.5.0 

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

SeriesPartitionCompactor represents an object reindexes a series partition and optionally compacts segments.

func NewSeriesPartitionCompactor added in v1.5.0 

func NewSeriesPartitionCompactor() *SeriesPartitionCompactor

NewSeriesPartitionCompactor returns a new instance of SeriesPartitionCompactor.

func (*SeriesPartitionCompactor) Compact added in v1.5.0 

func (c *SeriesPartitionCompactor) Compact(p *SeriesPartition) error

Compact rebuilds the series partition index.

type SeriesSegment added in v1.5.0 

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

SeriesSegment represents a log of series entries.

func CloneSeriesSegments added in v1.5.0 

func CloneSeriesSegments(a []*SeriesSegment) []*SeriesSegment

CloneSeriesSegments returns a copy of a slice of segments.

func CreateSeriesSegment added in v1.5.0 

func CreateSeriesSegment(id uint16, path string) (*SeriesSegment, error)

CreateSeriesSegment generates an empty segment at path.

func FindSegment added in v1.5.0 

func FindSegment(a []*SeriesSegment, id uint16) *SeriesSegment

FindSegment returns a segment by id.

func NewSeriesSegment added in v1.5.0 

func NewSeriesSegment(id uint16, path string) *SeriesSegment

NewSeriesSegment returns a new instance of SeriesSegment.

func (*SeriesSegment) AppendSeriesIDs added in v1.5.0 

func (s *SeriesSegment) AppendSeriesIDs(a []uint64) []uint64

AppendSeriesIDs appends all the segments ids to a slice. Returns the new slice.

func (*SeriesSegment) CanWrite added in v1.5.0 

func (s *SeriesSegment) CanWrite(data []byte) bool

CanWrite returns true if segment has space to write entry data.

func (*SeriesSegment) Clone added in v1.5.0 

func (s *SeriesSegment) Clone() *SeriesSegment

Clone returns a copy of the segment. Excludes the write handler, if set.

func (*SeriesSegment) Close added in v1.5.0 

func (s *SeriesSegment) Close() (err error)

Close unmaps the segment.

func (*SeriesSegment) CloseForWrite added in v1.5.0 

func (s *SeriesSegment) CloseForWrite() (err error)

func (*SeriesSegment) Data added in v1.6.0 

func (s *SeriesSegment) Data() []byte

Data returns the raw data.

func (*SeriesSegment) Flush added in v1.5.0 

func (s *SeriesSegment) Flush() error

Flush flushes the buffer to disk.

func (*SeriesSegment) ForEachEntry added in v1.5.0 

func (s *SeriesSegment) ForEachEntry(fn func(flag uint8, id uint64, offset int64, key []byte) error) error

ForEachEntry executes fn for every entry in the segment.

func (*SeriesSegment) ID added in v1.5.0 

func (s *SeriesSegment) ID() uint16

ID returns the id the segment was initialized with.

func (*SeriesSegment) InitForWrite added in v1.5.0 

func (s *SeriesSegment) InitForWrite() (err error)

InitForWrite initializes a write handle for the segment. This is only used for the last segment in the series file.

func (*SeriesSegment) MaxSeriesID added in v1.5.0 

func (s *SeriesSegment) MaxSeriesID() uint64

MaxSeriesID returns the highest series id in the segment.

func (*SeriesSegment) Open added in v1.5.0 

func (s *SeriesSegment) Open() error

Open memory maps the data file at the file's path.

func (*SeriesSegment) Size added in v1.5.0 

func (s *SeriesSegment) Size() int64

Size returns the size of the data in the segment. This is only populated once InitForWrite() is called.

func (*SeriesSegment) Slice added in v1.5.0 

func (s *SeriesSegment) Slice(pos uint32) []byte

Slice returns a byte slice starting at pos.

func (*SeriesSegment) WriteLogEntry added in v1.5.0 

func (s *SeriesSegment) WriteLogEntry(data []byte) (offset int64, err error)

WriteLogEntry writes entry data into the segment. Returns the offset of the beginning of the entry.

type SeriesSegmentHeader added in v1.5.0 

type SeriesSegmentHeader struct {
	Version uint8
}

SeriesSegmentHeader represents the header of a series segment.

func NewSeriesSegmentHeader added in v1.5.0 

func NewSeriesSegmentHeader() SeriesSegmentHeader

NewSeriesSegmentHeader returns a new instance of SeriesSegmentHeader.

func ReadSeriesSegmentHeader added in v1.5.0 

func ReadSeriesSegmentHeader(data []byte) (hdr SeriesSegmentHeader, err error)

ReadSeriesSegmentHeader returns the header from data.

func (*SeriesSegmentHeader) WriteTo added in v1.5.0 

func (hdr *SeriesSegmentHeader) WriteTo(w io.Writer) (n int64, err error)

WriteTo writes the header to w.

type Shard 

type Shard struct {
	EnableOnOpen bool

	// CompactionDisabled specifies the shard should not schedule compactions.
	// This option is intended for offline tooling.
	CompactionDisabled bool
	// contains filtered or unexported fields
}

Shard represents a self-contained time series database. An inverted index of the measurement and tag data is kept along with the raw time series data. Data can be split across many shards. The query engine in TSDB is responsible for combining the output of many shards into a single query result.

func NewShard 

func NewShard(id uint64, path string, walPath string, sfile *SeriesFile, opt EngineOptions) *Shard

NewShard returns a new initialized Shard. walPath doesn't apply to the b1 type index

func (*Shard) Backup added in v1.3.6 

func (s *Shard) Backup(w io.Writer, basePath string, since time.Time) error

Backup backs up the shard by creating a tar archive of all TSM files that have been modified since the provided time. See Engine.Backup for more details.

func (*Shard) Close 

func (s *Shard) Close() error

Close shuts down the shard's store.

func (*Shard) CreateCursorIterator added in v1.6.0 

func (s *Shard) CreateCursorIterator(ctx context.Context) (CursorIterator, error)

func (*Shard) CreateIterator added in v0.11.0 

func (s *Shard) CreateIterator(ctx context.Context, m *influxql.Measurement, opt query.IteratorOptions) (query.Iterator, error)

CreateIterator returns an iterator for the data in the shard.

func (*Shard) CreateSeriesCursor added in v1.6.0 

func (s *Shard) CreateSeriesCursor(ctx context.Context, req SeriesCursorRequest, cond influxql.Expr) (SeriesCursor, error)

func (*Shard) CreateSnapshot added in v1.0.0 

func (s *Shard) CreateSnapshot() (string, error)

CreateSnapshot will return a path to a temp directory containing hard links to the underlying shard files.

func (*Shard) Database added in v1.3.0 

func (s *Shard) Database() string

Database returns the database of the shard.

func (*Shard) DeleteMeasurement added in v0.9.3 

func (s *Shard) DeleteMeasurement(name []byte) error

DeleteMeasurement deletes a measurement and all underlying series.

func (*Shard) DeleteSeriesRange added in v0.13.0 

func (s *Shard) DeleteSeriesRange(itr SeriesIterator, min, max int64) error

DeleteSeriesRange deletes all values from for seriesKeys between min and max (inclusive)

func (*Shard) DeleteSeriesRangeWithPredicate added in v1.6.0 

func (s *Shard) DeleteSeriesRangeWithPredicate(itr SeriesIterator, predicate func(name []byte, tags models.Tags) (int64, int64, bool)) error

DeleteSeriesRangeWithPredicate deletes all values from for seriesKeys between min and max (inclusive) for which predicate() returns true. If predicate() is nil, then all values in range are deleted.

func (*Shard) Digest added in v1.5.0 

func (s *Shard) Digest() (io.ReadCloser, int64, error)

Digest returns a digest of the shard.

func (*Shard) DiskSize added in v0.9.4 

func (s *Shard) DiskSize() (int64, error)

DiskSize returns the size on disk of this shard.

func (*Shard) Engine added in v1.5.5 

func (s *Shard) Engine() (Engine, error)

engine safely (under an RLock) returns a reference to the shard's Engine, or an error if the Engine is closed, or the shard is currently disabled.

The shard's Engine should always be accessed via a call to engine(), rather than directly referencing Shard.engine.

If a caller needs an Engine reference but is already under a lock, then they should use engineNoLock().

func (*Shard) Export added in v1.5.0 

func (s *Shard) Export(w io.Writer, basePath string, start time.Time, end time.Time) error

func (*Shard) FieldDimensions added in v0.11.0 

func (s *Shard) FieldDimensions(measurements []string) (fields map[string]influxql.DataType, dimensions map[string]struct{}, err error)

FieldDimensions returns unique sets of fields and dimensions across a list of sources.

func (*Shard) ForEachMeasurementName added in v1.3.6 

func (s *Shard) ForEachMeasurementName(fn func(name []byte) error) error

ForEachMeasurementName iterates over each measurement in the shard.

func (*Shard) Free added in v1.4.0 

func (s *Shard) Free() error

func (*Shard) ID added in v1.3.0 

func (s *Shard) ID() uint64

ID returns the shards ID.

func (*Shard) Import added in v1.3.0 

func (s *Shard) Import(r io.Reader, basePath string) error

Import imports data to the underlying engine for the shard. r should be a reader from a backup created by Backup.

func (*Shard) Index added in v1.4.0 

func (s *Shard) Index() (Index, error)

Index returns a reference to the underlying index. It returns an error if the index is nil.

func (*Shard) IndexType added in v1.3.0 

func (s *Shard) IndexType() string

IndexType returns the index version being used for this shard.

IndexType returns the empty string if it is called before the shard is opened, since it is only that point that the underlying index type is known.

func (*Shard) IsIdle added in v1.3.0 

func (s *Shard) IsIdle() bool

IsIdle return true if the shard is not receiving writes and is fully compacted.

func (*Shard) LastModified added in v1.2.0 

func (s *Shard) LastModified() time.Time

LastModified returns the time when this shard was last modified.

func (*Shard) MeasurementExists added in v1.3.0 

func (s *Shard) MeasurementExists(name []byte) (bool, error)

MeasurementExists returns true if the shard contains name. TODO(edd): This method is currently only being called from tests; do we really need it?

func (*Shard) MeasurementFields added in v1.3.0 

func (s *Shard) MeasurementFields(name []byte) *MeasurementFields

MeasurementFields returns fields for a measurement. TODO(edd): This method is currently only being called from tests; do we really need it?

func (*Shard) MeasurementNamesByRegex added in v1.3.6 

func (s *Shard) MeasurementNamesByRegex(re *regexp.Regexp) ([][]byte, error)

MeasurementNamesByRegex returns names of measurements matching the regular expression.

func (*Shard) MeasurementTagKeyValuesByExpr added in v1.3.6 

func (s *Shard) MeasurementTagKeyValuesByExpr(auth query.Authorizer, name []byte, key []string, expr influxql.Expr, keysSorted bool) ([][]string, error)

MeasurementTagKeyValuesByExpr returns all the tag keys values for the provided expression.

func (*Shard) MeasurementTagKeysByExpr added in v1.3.6 

func (s *Shard) MeasurementTagKeysByExpr(name []byte, expr influxql.Expr) (map[string]struct{}, error)

MeasurementTagKeysByExpr returns all the tag keys for the provided expression.

func (*Shard) MeasurementsSketches added in v1.3.0 

func (s *Shard) MeasurementsSketches() (estimator.Sketch, estimator.Sketch, error)

MeasurementsSketches returns the measurement sketches for the shard.

func (*Shard) Open 

func (s *Shard) Open() error

Open initializes and opens the shard's store.

func (*Shard) Path 

func (s *Shard) Path() string

Path returns the path set on the shard when it was created.

func (*Shard) Restore added in v1.0.0 

func (s *Shard) Restore(r io.Reader, basePath string) error

Restore restores data to the underlying engine for the shard. The shard is reopened after restore.

func (*Shard) RetentionPolicy added in v1.3.0 

func (s *Shard) RetentionPolicy() string

RetentionPolicy returns the retention policy of the shard.

func (*Shard) ScheduleFullCompaction added in v1.5.0 

func (s *Shard) ScheduleFullCompaction() error

ScheduleFullCompaction forces a full compaction to be schedule on the shard.

func (*Shard) SeriesFile added in v1.5.5 

func (s *Shard) SeriesFile() (*SeriesFile, error)

SeriesFile returns a reference the underlying series file. If return an error if the series file is nil.

func (*Shard) SeriesN added in v1.3.0 

func (s *Shard) SeriesN() int64

SeriesN returns the unique number of series in the shard.

func (*Shard) SeriesSketches added in v1.3.0 

func (s *Shard) SeriesSketches() (estimator.Sketch, estimator.Sketch, error)

SeriesSketches returns the measurement sketches for the shard.

func (*Shard) SetCompactionsEnabled added in v1.3.0 

func (s *Shard) SetCompactionsEnabled(enabled bool)

SetCompactionsEnabled enables or disable shard background compactions.

func (*Shard) SetEnabled added in v1.0.0 

func (s *Shard) SetEnabled(enabled bool)

SetEnabled enables the shard for queries and write. When disabled, all writes and queries return an error and compactions are stopped for the shard.

func (*Shard) Statistics added in v1.0.0 

func (s *Shard) Statistics(tags map[string]string) []models.Statistic

Statistics returns statistics for periodic monitoring.

func (*Shard) TagKeyCardinality added in v1.3.0 

func (s *Shard) TagKeyCardinality(name, key []byte) int

func (*Shard) WithLogger added in v1.2.0 

func (s *Shard) WithLogger(log *zap.Logger)

WithLogger sets the logger on the shard. It must be called before Open.

func (*Shard) WritePoints 

func (s *Shard) WritePoints(points []models.Point) error

WritePoints will write the raw data points and any new metadata to the index in the shard.

func (*Shard) WriteTo added in v0.9.4 

func (s *Shard) WriteTo(w io.Writer) (int64, error)

WriteTo writes the shard's data to w.

type ShardError added in v0.11.0 

type ShardError struct {
	Err error
	// contains filtered or unexported fields
}

A ShardError implements the error interface, and contains extra context about the shard that generated the error.

func (ShardError) Error added in v0.11.0 

func (e ShardError) Error() string

Error returns the string representation of the error, to satisfy the error interface.

type ShardGroup added in v1.2.0 

type ShardGroup interface {
	MeasurementsByRegex(re *regexp.Regexp) []string
	FieldKeysByMeasurement(name []byte) []string
	FieldDimensions(measurements []string) (fields map[string]influxql.DataType, dimensions map[string]struct{}, err error)
	MapType(measurement, field string) influxql.DataType
	CreateIterator(ctx context.Context, measurement *influxql.Measurement, opt query.IteratorOptions) (query.Iterator, error)
	IteratorCost(measurement string, opt query.IteratorOptions) (query.IteratorCost, error)
	ExpandSources(sources influxql.Sources) (influxql.Sources, error)
}

type ShardStatistics added in v1.0.0 

type ShardStatistics struct {
	WriteReq           int64
	WriteReqOK         int64
	WriteReqErr        int64
	FieldsCreated      int64
	WritePointsErr     int64
	WritePointsDropped int64
	WritePointsOK      int64
	BytesWritten       int64
	DiskBytes          int64
}

ShardStatistics maintains statistics for a shard.

type Shards added in v0.11.0 

type Shards []*Shard

Shards represents a sortable list of shards.

func (Shards) CallType added in v1.6.0 

func (a Shards) CallType(name string, args []influxql.DataType) (influxql.DataType, error)

func (Shards) CreateIterator added in v1.2.0 

func (a Shards) CreateIterator(ctx context.Context, measurement *influxql.Measurement, opt query.IteratorOptions) (query.Iterator, error)

func (Shards) CreateSeriesCursor added in v1.6.0 

func (a Shards) CreateSeriesCursor(ctx context.Context, req SeriesCursorRequest, cond influxql.Expr) (_ SeriesCursor, err error)

func (Shards) ExpandSources added in v1.2.0 

func (a Shards) ExpandSources(sources influxql.Sources) (influxql.Sources, error)

func (Shards) FieldDimensions added in v1.2.0 

func (a Shards) FieldDimensions(measurements []string) (fields map[string]influxql.DataType, dimensions map[string]struct{}, err error)

func (Shards) FieldKeysByMeasurement added in v1.6.1 

func (a Shards) FieldKeysByMeasurement(name []byte) []string

FieldKeysByMeasurement returns a de-duplicated, sorted, set of field keys for the provided measurement name.

func (Shards) IteratorCost added in v1.4.0 

func (a Shards) IteratorCost(measurement string, opt query.IteratorOptions) (query.IteratorCost, error)

func (Shards) Len added in v0.11.0 

func (a Shards) Len() int

Len implements sort.Interface.

func (Shards) Less added in v0.11.0 

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

Less implements sort.Interface.

func (Shards) MapType added in v1.2.0 

func (a Shards) MapType(measurement, field string) influxql.DataType

func (Shards) MeasurementsByRegex added in v1.2.0 

func (a Shards) MeasurementsByRegex(re *regexp.Regexp) []string

MeasurementsByRegex returns the unique set of measurements matching the provided regex, for all the shards.

func (Shards) Swap added in v0.11.0 

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

Swap implements sort.Interface.

type Store 

type Store struct {
	SeriesFileMaxSize int64 // Determines size of series file mmap. Can be altered in tests.

	EngineOptions EngineOptions

	Logger *zap.Logger
	// contains filtered or unexported fields
}

Store manages shards and indexes for databases.

func NewStore 

func NewStore(path string) *Store

NewStore returns a new store with the given path and a default configuration. The returned store must be initialized by calling Open before using it.

func (*Store) BackupShard added in v0.10.0 

func (s *Store) BackupShard(id uint64, since time.Time, w io.Writer) error

BackupShard will get the shard and have the engine backup since the passed in time to the writer.

func (*Store) Close 

func (s *Store) Close() error

Close closes the store and all associated shards. After calling Close accessing shards through the Store will result in ErrStoreClosed being returned.

func (*Store) CreateShard 

func (s *Store) CreateShard(database, retentionPolicy string, shardID uint64, enabled bool) error

CreateShard creates a shard with the given id and retention policy on a database.

func (*Store) CreateShardSnapshot added in v1.0.0 

func (s *Store) CreateShardSnapshot(id uint64) (string, error)

CreateShardSnapShot will create a hard link to the underlying shard and return a path. The caller is responsible for cleaning up (removing) the file path returned.

func (*Store) Databases added in v0.9.4 

func (s *Store) Databases() []string

Databases returns the names of all databases managed by the store.

func (*Store) DeleteDatabase 

func (s *Store) DeleteDatabase(name string) error

DeleteDatabase will close all shards associated with a database and remove the directory and files from disk.

func (*Store) DeleteMeasurement added in v0.11.0 

func (s *Store) DeleteMeasurement(database, name string) error

DeleteMeasurement removes a measurement and all associated series from a database.

func (*Store) DeleteRetentionPolicy added in v0.11.0 

func (s *Store) DeleteRetentionPolicy(database, name string) error

DeleteRetentionPolicy will close all shards associated with the provided retention policy, remove the retention policy directories on both the DB and WAL, and remove all shard files from disk.

func (*Store) DeleteSeries added in v0.11.0 

func (s *Store) DeleteSeries(database string, sources []influxql.Source, condition influxql.Expr) error

DeleteSeries loops through the local shards and deletes the series data for the passed in series keys.

func (*Store) DeleteShard 

func (s *Store) DeleteShard(shardID uint64) error

DeleteShard removes a shard from disk.

func (*Store) DiskSize added in v0.9.4 

func (s *Store) DiskSize() (int64, error)

DiskSize returns the size of all the shard files in bytes. This size does not include the WAL size.

func (*Store) ExpandSources added in v0.11.0 

func (s *Store) ExpandSources(sources influxql.Sources) (influxql.Sources, error)

ExpandSources expands sources against all local shards.

func (*Store) ExportShard added in v1.5.0 

func (s *Store) ExportShard(id uint64, start time.Time, end time.Time, w io.Writer) error

func (*Store) ImportShard added in v1.3.0 

func (s *Store) ImportShard(id uint64, r io.Reader) error

ImportShard imports the contents of r to a given shard. All files in the backup are added as new files which may cause duplicated data to occur requiring more expensive compactions.

func (*Store) IndexBytes added in v1.6.0 

func (s *Store) IndexBytes() int

func (*Store) MeasurementNames added in v1.3.0 

func (s *Store) MeasurementNames(auth query.Authorizer, database string, cond influxql.Expr) ([][]byte, error)

MeasurementNames returns a slice of all measurements. Measurements accepts an optional condition expression. If cond is nil, then all measurements for the database will be returned.

func (*Store) MeasurementSeriesCounts added in v1.3.0 

func (s *Store) MeasurementSeriesCounts(database string) (measuments int, series int)

MeasurementSeriesCounts returns the number of measurements and series in all the shards' indices.

func (*Store) MeasurementsCardinality added in v1.3.0 

func (s *Store) MeasurementsCardinality(database string) (int64, error)

MeasurementsCardinality returns an estimation of the measurement cardinality for the provided database.

Cardinality is calculated using a sketch-based estimation. The result of this method cannot be combined with any other results.

func (*Store) MeasurementsSketches added in v1.5.0 

func (s *Store) MeasurementsSketches(database string) (estimator.Sketch, estimator.Sketch, error)

MeasurementsSketches returns the sketches associated with the measurement data in all the shards in the provided database.

The returned sketches can be combined with other sketches to provide an estimation across distributed databases.

func (*Store) Open 

func (s *Store) Open() error

Open initializes the store, creating all necessary directories, loading all shards as well as initializing periodic maintenance of them.

func (*Store) Path 

func (s *Store) Path() string

Path returns the store's root path.

func (*Store) RestoreShard added in v1.0.0 

func (s *Store) RestoreShard(id uint64, r io.Reader) error

RestoreShard restores a backup from r to a given shard. This will only overwrite files included in the backup.

func (*Store) SeriesCardinality added in v1.3.0 

func (s *Store) SeriesCardinality(database string) (int64, error)

SeriesCardinality returns the exact series cardinality for the provided database.

Cardinality is calculated exactly by unioning all shards' bitsets of series IDs. The result of this method cannot be combined with any other results.

func (*Store) SeriesSketches added in v1.5.0 

func (s *Store) SeriesSketches(database string) (estimator.Sketch, estimator.Sketch, error)

SeriesSketches returns the sketches associated with the series data in all the shards in the provided database.

The returned sketches can be combined with other sketches to provide an estimation across distributed databases.

func (*Store) SetShardEnabled added in v1.0.0 

func (s *Store) SetShardEnabled(shardID uint64, enabled bool) error

SetShardEnabled enables or disables a shard for read and writes.

func (*Store) Shard 

func (s *Store) Shard(id uint64) *Shard

Shard returns a shard by id.

func (*Store) ShardDigest added in v1.5.0 

func (s *Store) ShardDigest(id uint64) (io.ReadCloser, int64, error)

ShardDigest returns a digest of the shard with the specified ID.

func (*Store) ShardGroup added in v1.2.0 

func (s *Store) ShardGroup(ids []uint64) ShardGroup

ShardGroup returns a ShardGroup with a list of shards by id.

func (*Store) ShardIDs 

func (s *Store) ShardIDs() []uint64

ShardIDs returns a slice of all ShardIDs under management.

func (*Store) ShardN added in v0.9.3 

func (s *Store) ShardN() int

ShardN returns the number of shards in the store.

func (*Store) ShardRelativePath added in v0.10.0 

func (s *Store) ShardRelativePath(id uint64) (string, error)

ShardRelativePath will return the relative path to the shard, i.e., <database>/<retention>/<id>.

func (*Store) Shards added in v0.11.0 

func (s *Store) Shards(ids []uint64) []*Shard

Shards returns a list of shards by id.

func (*Store) Statistics added in v1.0.0 

func (s *Store) Statistics(tags map[string]string) []models.Statistic

Statistics returns statistics for period monitoring.

func (*Store) TagKeys added in v1.4.0 

func (s *Store) TagKeys(auth query.Authorizer, shardIDs []uint64, cond influxql.Expr) ([]TagKeys, error)

TagKeys returns the tag keys in the given database, matching the condition.

func (*Store) TagValues added in v1.0.0 

func (s *Store) TagValues(auth query.Authorizer, shardIDs []uint64, cond influxql.Expr) ([]TagValues, error)

TagValues returns the tag keys and values for the provided shards, where the tag values satisfy the provided condition.

func (*Store) WithLogger added in v1.2.0 

func (s *Store) WithLogger(log *zap.Logger)

WithLogger sets the logger for the store.

func (*Store) WriteToShard 

func (s *Store) WriteToShard(shardID uint64, points []models.Point) error

WriteToShard writes a list of points to a shard identified by its ID.

type StringArray added in v1.7.0 

type StringArray = cursors.StringArray

func NewStringArrayLen added in v1.7.0 

func NewStringArrayLen(sz int) *StringArray

type StringArrayCursor added in v1.7.0 

type StringArrayCursor = cursors.StringArrayCursor

type TagKeyIterator added in v1.5.0 

type TagKeyIterator interface {
	Close() error
	Next() ([]byte, error)
}

TagKeyIterator represents a iterator over a list of tag keys.

func MergeTagKeyIterators added in v1.5.0 

func MergeTagKeyIterators(itrs ...TagKeyIterator) TagKeyIterator

MergeTagKeyIterators returns an iterator that merges a set of iterators.

type TagKeyIterators added in v1.5.0 

type TagKeyIterators []TagKeyIterator

func (TagKeyIterators) Close added in v1.5.0 

func (a TagKeyIterators) Close() (err error)

type TagKeys added in v1.4.0 

type TagKeys struct {
	Measurement string
	Keys        []string
}

type TagKeysSlice added in v1.4.0 

type TagKeysSlice []TagKeys

func (TagKeysSlice) Len added in v1.4.0 

func (a TagKeysSlice) Len() int

func (TagKeysSlice) Less added in v1.4.0 

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

func (TagKeysSlice) Swap added in v1.4.0 

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

type TagValueIterator added in v1.5.0 

type TagValueIterator interface {
	Close() error
	Next() ([]byte, error)
}

TagValueIterator represents a iterator over a list of tag values.

func MergeTagValueIterators added in v1.5.0 

func MergeTagValueIterators(itrs ...TagValueIterator) TagValueIterator

MergeTagValueIterators returns an iterator that merges a set of iterators.

type TagValueIterators added in v1.5.0 

type TagValueIterators []TagValueIterator

func (TagValueIterators) Close added in v1.5.0 

func (a TagValueIterators) Close() (err error)

type TagValues added in v1.0.0 

type TagValues struct {
	Measurement string
	Values      []KeyValue
}

type TagValuesSlice added in v1.3.0 

type TagValuesSlice []TagValues

func (TagValuesSlice) Len added in v1.3.0 

func (a TagValuesSlice) Len() int

func (TagValuesSlice) Less added in v1.3.0 

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

func (TagValuesSlice) Swap added in v1.3.0 

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

type UnsignedArray added in v1.7.0 

type UnsignedArray = cursors.UnsignedArray

func NewUnsignedArrayLen added in v1.7.0 

func NewUnsignedArrayLen(sz int) *UnsignedArray

type UnsignedArrayCursor added in v1.7.0 

type UnsignedArrayCursor = cursors.UnsignedArrayCursor

Source Files

View all Source files

Directories

Path Synopsis
Package engine can be imported to initialize and register all available TSDB engines.
 Package engine can be imported to initialize and register all available TSDB engines.
tsm1
Package tsm1 provides a TSDB in the Time Structured Merge tree format.
 Package tsm1 provides a TSDB in the Time Structured Merge tree format.
inmem
Package inmem implements a shared, in-memory index for each database.
 Package inmem implements a shared, in-memory index for each database.
internal
tsi1
Package tsi1 provides a memory-mapped index implementation that supports high cardinality series.
 Package tsi1 provides a memory-mapped index implementation that supports high cardinality series.
Package tsdb is a generated protocol buffer package.
 Package tsdb is a generated protocol buffer package.

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.