This section provides information about the different window types available in Macrometa stream workers.
This window outputs the arriving events as and when they arrive and resets (expires) the window periodically based on the given cron expression.
A delay window holds events for a specific time period that is regarded as a delay period before processing them.
Holds events that belong to a session. Events that belong to a specific session are identified by a session key, and a session gap is determines the time period after which the session is considered to be expired. To have meaningful aggregation on session windows, the events need to be aggregated based on the session key via a group by clause.
A sliding window that holds the last number of events defined by an expression at a given time, and gets updated for each arrival and expiration.
A sliding time window based on external time. It holds events that arrived during the last window window.time period from the external timestamp, and gets updated on every monotonically increasing timestamp.
A sliding length window that holds the last length events at a given time, and gets updated for each arrival and expiration.
A sliding time window that at a given time holds the last length events that arrived during last time period, and gets updated for every event arrival and expiration.
A sliding time window that holds events that arrived during the last window time period at a given time, and gets updated for each event arrival and expiration.
This window holds a batch of events that equal the number specified as the window length and sorts them in the given order.
A batch (tumbling) window that holds and process a number of events as specified in the expression.
A batch (tumbling) time window based on external time, that holds events arrived during window.time periods, and gets updated for every window.time.
A batch (tumbling) length window that holds and process a number of events as specified in the length.
A tumbling time batch window holds and processes events that arrive during the time period as a batch.
A window that holds an incoming events batch. When a new set of events arrives, the old events are expired. Tumbling windows can be used to aggregate events that comes in batches. If it has the parameter length specified, then the batch window processes the batch as several chunks.