Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Dispatcher ¶
type Dispatcher interface {
// Start initializes and runs the dispatcher based on different implementations.
Start(ctx context.Context)
// Dispatch determines if the current Maestro instance should process the resource status update based on the consumer ID.
Dispatch(consumerName string) bool
// OnInstanceUp is called when a new maestro instance is up.
OnInstanceUp(instanceID string) error
// OnInstanceDown is called when a maestro instance is inactive.
OnInstanceDown(instanceID string) error
}
Dispatcher defines methods for coordinating resource status updates in the context of multiple active maestro instances.
The dispatcher ensures only one instance processes specific resource status updates from a consumer. It needs to handle status resync based on the instances' status and different implementations.
type HashDispatcher ¶
type HashDispatcher struct {
// contains filtered or unexported fields
}
HashDispatcher is an implementation of Dispatcher. It uses consistent hashing to map consumers to maestro instances. Only the maestro instance that is mapped to a consumer will process the resource status update from that consumer. Need to trigger status resync for the consumer when an instance is up or down.
func NewHashDispatcher ¶
func NewHashDispatcher(instanceID string, instanceDao dao.InstanceDao, consumerDao dao.ConsumerDao, sourceClient cloudevents.SourceClient) *HashDispatcher
func (*HashDispatcher) Dispatch ¶
func (d *HashDispatcher) Dispatch(consumerName string) bool
Dispatch checks if the provided consumer ID is owned by the current maestro instance. It returns true if the consumer is part of the current instance's consumer set; otherwise, it returns false.
func (*HashDispatcher) OnInstanceDown ¶
func (d *HashDispatcher) OnInstanceDown(instanceID string) error
OnInstanceDown removes the instance from the hashing ring and updates the consumer set for the current instance.
func (*HashDispatcher) OnInstanceUp ¶
func (d *HashDispatcher) OnInstanceUp(instanceID string) error
OnInstanceUp adds the new instance to the hashing ring and updates the consumer set for the current instance.
func (*HashDispatcher) Start ¶
func (d *HashDispatcher) Start(ctx context.Context)
Start initializes and runs the dispatcher, updating the hashing ring and consumer set for the current instance.
type NoopDispatcher ¶
type NoopDispatcher struct {
// contains filtered or unexported fields
}
NoopDispatcher is a no-op implementation of Dispatcher. It will always dispatch the resource status update to the current maestro instance. This is the default implementation when shared subscription is enabled. Need to trigger status resync from all consumers when an instance is down.
func NewNoopDispatcher ¶
func NewNoopDispatcher(consumerDao dao.ConsumerDao, sourceClient cloudevents.SourceClient) *NoopDispatcher
NewNoopDispatcher creates a new NoopDispatcher instance.
func (*NoopDispatcher) Dispatch ¶
func (d *NoopDispatcher) Dispatch(consumerID string) bool
Dispatch always returns true, indicating that the current maestro instance should process the resource status update.
func (*NoopDispatcher) OnInstanceDown ¶
func (d *NoopDispatcher) OnInstanceDown(instanceID string) error
OnInstanceDown triggers status resync from all consumers.
func (*NoopDispatcher) OnInstanceUp ¶
func (d *NoopDispatcher) OnInstanceUp(instanceID string) error
OnInstanceUp is a no-op implementation.
func (*NoopDispatcher) Start ¶
func (d *NoopDispatcher) Start(ctx context.Context)
Start is a no-op implementation.
Source Files