> ## Documentation Index > Fetch the complete documentation index at: https://docs.prefect.io/llms.txt > Use this file to discover all available pages before exploring further. # memory # `prefect.server.events.ordering.memory` ## Classes ### `EventBeingProcessed` Indicates that an event is currently being processed and should not be processed until it is finished. This may happen due to concurrent processing. ### `CausalOrdering` **Methods:** #### `clear` ```python theme={null} clear(self) -> None ``` Clear all data for this scope. #### `clear_all_scopes` ```python theme={null} clear_all_scopes(cls) -> None ``` Clear all data for all scopes - useful for testing. #### `event_has_been_seen` ```python theme={null} event_has_been_seen(self, event: UUID | Event) -> bool ``` #### `event_has_started_processing` ```python theme={null} event_has_started_processing(self, event: UUID | Event) -> bool ``` #### `event_is_processing` ```python theme={null} event_is_processing(self, event: ReceivedEvent) -> AsyncGenerator[None, None] ``` Mark an event as being processed for the duration of its lifespan through the ordering system. #### `followers_by_id` ```python theme={null} followers_by_id(self, follower_ids: list[UUID]) -> list[ReceivedEvent] ``` Returns the events with the given IDs, in the order they occurred. #### `forget_event_is_processing` ```python theme={null} forget_event_is_processing(self, event: ReceivedEvent) -> None ``` #### `forget_follower` ```python theme={null} forget_follower(self, follower: ReceivedEvent) -> None ``` Forget that this event is waiting on another event to arrive. #### `get_followers` ```python theme={null} get_followers(self, leader: ReceivedEvent) -> list[ReceivedEvent] ``` Returns events that were waiting on this leader event to arrive. #### `get_lost_followers` ```python theme={null} get_lost_followers(self) -> list[ReceivedEvent] ``` Returns events that were waiting on a leader event that never arrived. #### `preceding_event_confirmed` ```python theme={null} preceding_event_confirmed(self, handler: event_handler, event: ReceivedEvent, depth: int = 0) -> AsyncGenerator[None, None] ``` Events may optionally declare that they logically follow another event, so that we can preserve important event orderings in the face of unreliable delivery and ordering of messages from the queues. This function keeps track of the ID of each event that this shard has successfully processed going back to the PRECEDING\_EVENT\_LOOKBACK period. If an event arrives that must follow another one, confirm that we have recently seen and processed that event before proceeding. **Args:** * `handler`: The function to call when an out-of-order event is ready to be processed * `event`: The event to be processed. This object should include metadata indicating if and what event it follows. * `depth`: The current recursion depth, used to prevent infinite recursion due to cyclic dependencies between events. Defaults to 0. Raises EventArrivedEarly if the current event shouldn't be processed yet. #### `record_event_as_processing` ```python theme={null} record_event_as_processing(self, event: ReceivedEvent) -> bool ``` Record that an event is being processed, returning False if already processing. #### `record_event_as_seen` ```python theme={null} record_event_as_seen(self, event: ReceivedEvent) -> None ``` #### `record_follower` ```python theme={null} record_follower(self, event: ReceivedEvent) -> None ``` Remember that this event is waiting on another event to arrive. #### `wait_for_leader` ```python theme={null} wait_for_leader(self, event: ReceivedEvent) -> None ``` Given an event, wait for its leader to be processed before proceeding, or raise EventArrivedEarly if we would wait too long in this attempt.