catena

README

catena

n. A closely linked series.

Catena is a storage engine for time series data.

Documentation

Extensive documentation is not available yet, but there are a few resources you can use to learn more:

1. Original blog post introduction Catena: http://misfra.me/state-of-the-state-part-iii
2. Time series storage slide deck: https://speakerdeck.com/preetamjinka/time-series-storage

Disclaimer

Catena is just a prototype at the moment. It's not ready for you to use.

License

BSD (see LICENSE)

Documentation

Overview

Package catena provides a time series storage engine.

Index

• type DB
  • func NewDB(baseDir string, partitionSize, maxPartitions int) (*DB, error)
  • func OpenDB(baseDir string, partitionSize, maxPartitions int) (*DB, error)
  • func (db *DB) Close() error
  • func (db *DB) InsertRows(rows []Row) error
  • func (db *DB) Metrics(source string, start, end int64) []string
  • func (db *DB) NewIterator(source, metric string) (*Iterator, error)
  • func (db *DB) Sources(start, end int64) []string
• type Iterator
  • func (i *Iterator) Close()
  • func (i *Iterator) Next() error
  • func (i *Iterator) Reset() error
  • func (i *Iterator) Seek(timestamp int64) error
• type Point
• type Row

Types

type DB

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

DB is a handle to a catena database.

func NewDB

func NewDB(baseDir string, partitionSize, maxPartitions int) (*DB, error)

NewDB creates a new DB located in baseDir. If baseDir does not exist it will be created. An error is returned if baseDir is not empty.

func OpenDB

func OpenDB(baseDir string, partitionSize, maxPartitions int) (*DB, error)

OpenDB opens a DB located in baseDir.

func (*DB) Close

func (db *DB) Close() error

Close closes the DB and releases any internal state. Close will block if there are active iterators.

func (*DB) InsertRows

func (db *DB) InsertRows(rows []Row) error

InsertRows inserts the given rows into the database.

func (*DB) Metrics

func (db *DB) Metrics(source string, start, end int64) []string

Metrics returns a slice of metrics that are present within the given time range for the given source.

func (*DB) NewIterator

func (db *DB) NewIterator(source, metric string) (*Iterator, error)

NewIterator creates a new Iterator for the given source and metric.

func (*DB) Sources

func (db *DB) Sources(start, end int64) []string

Sources returns a slice of sources that are present within the given time range.

type Iterator

type Iterator struct {
	partition.Iterator
	// contains filtered or unexported fields
}

An Iterator is a cursor over an array of points for a source and metric.

func (*Iterator) Close

func (i *Iterator) Close()

Close closes the iterator. Iterators MUST be closed to unblock the compactor!

func (*Iterator) Next

func (i *Iterator) Next() error

Next advances i to the next available point.

func (*Iterator) Reset

func (i *Iterator) Reset() error

Reset moves i to the first available timestamp.

func (*Iterator) Seek

func (i *Iterator) Seek(timestamp int64) error

Seek moves the iterator to the first timestamp greater than or equal to timestamp.

type Point

type Point struct {
	Timestamp int64   `json:"timestamp"`
	Value     float64 `json:"value"`
}

A Point is a single observation of a time series metric. It has a timestamp and a value.

type Row

type Row struct {
	Source string `json:"source"`
	Metric string `json:"metric"`
	Point
}

A Row is a Point with Source and Metric fields.