Documentation
¶
Overview ¶
Package bigsegments contains logic for synchronizing big segments.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type BigSegmentStore ¶
type BigSegmentStore interface {
io.Closer
// GetSynchronizedOn returns the synchronization time from the external store.
//
// The synchronization time may not exist in the store. Use `IsDefined()` to
// check the result.
GetSynchronizedOn() (ldtime.UnixMillisecondTime, error)
// contains filtered or unexported methods
}
BigSegmentStore is the interface for interacting with an external big segment store. Each instance is specific to one LD environment.
func DefaultBigSegmentStoreFactory ¶
func DefaultBigSegmentStoreFactory( envConfig config.EnvConfig, allConfig config.Config, loggers ldlog.Loggers, ) (BigSegmentStore, error)
DefaultBigSegmentStoreFactory implements our standard logic for optionally creating a BigSegmentStore.
func NewNullBigSegmentStore ¶
func NewNullBigSegmentStore() BigSegmentStore
NewNullBigSegmentStore returns a no-op stub implementation. This is used only in tests, but it is exported from this package so that we can keep the interface methods private.
type BigSegmentStoreFactory ¶
type BigSegmentStoreFactory func( envConfig config.EnvConfig, allConfig config.Config, loggers ldlog.Loggers, ) (BigSegmentStore, error)
BigSegmentStoreFactory creates an implementation of BigSegmentStore, if the configuration implies that we should have one; if not, it returns nil.
type BigSegmentSynchronizer ¶
type BigSegmentSynchronizer interface {
// Start begins synchronization of an environment.
//
// This method does not block.
//
// If the BigSegmentSynchronizer has already been started, or has been closed, this has no effect.
Start()
// HasSynced returns true if the synchronizer has ever successfully synced the data.
//
// We use this to determine whether Relay's internal SDK instances should bother trying to query
// big segments metadata. If we haven't yet written any metadata, then trying to do so would
// produce useless errors.
HasSynced() bool
// SegmentUpdatesCh returns a channel for notifications about segment data updates.
//
// Each value posted to this channel represents a batch of updates that the synchronizer has
// applied. The caller is responsible for reading the channel to avoid blocking the
// synchronizer.
SegmentUpdatesCh() <-chan UpdatesSummary
// Close ends synchronization of an environment.
//
// This method does not block.
//
// The BigSegmentSynchronizer cannot be restarted after calling Close.
Close()
}
BigSegmentSynchronizer synchronizes big segment state for a given environment.
func DefaultBigSegmentSynchronizerFactory ¶
func DefaultBigSegmentSynchronizerFactory( httpConfig httpconfig.HTTPConfig, store BigSegmentStore, pollURI string, streamURI string, envID config.EnvironmentID, sdkKey config.SDKKey, loggers ldlog.Loggers, logPrefix string, ) BigSegmentSynchronizer
DefaultBigSegmentSynchronizerFactory creates the default implementation of BigSegmentSynchronizer.
type BigSegmentSynchronizerFactory ¶
type BigSegmentSynchronizerFactory func( httpConfig httpconfig.HTTPConfig, store BigSegmentStore, pollURI string, streamURI string, envID config.EnvironmentID, sdkKey config.SDKKey, loggers ldlog.Loggers, logPrefix string, ) BigSegmentSynchronizer
BigSegmentSynchronizerFactory creates an implementation of BigSegmentSynchronizer. We only use a single implementation in real life, but this allows us to use a mock one in tests. Calling the factory does not automatically start the synchronizer.
type UpdatesSummary ¶
type UpdatesSummary struct {
// SegmentKeysUpdated is a slice of segment keys (plain keys as used by the SDK-- not segment
// IDs, i.e. there is no generation suffix).
SegmentKeysUpdated []string
}
UpdatesSummary describes a batch of updates that the synchronizer has applied.
Source Files