

API Docs - v4.0.6

Kslack

reorder (Stream Processor)

This stream processor extension performs reordering of an out-of-order event stream.
 It implements the K-Slack based out-of-order handling algorithm (originally described in
https://www2.informatik.uni-erlangen.de/publication/download/IPDPS2013.pdf)

Syntax

kslack:reorder(<LONG> timestamp, <LONG> timer.timeout, <LONG> max.k, <BOOL> discard.flag)

QUERY PARAMETERS

Name	Description	Default Value	Possible Data Types	Optional	Dynamic
timestamp	Attribute used for used for ordering the events		LONG	No	No
timer.timeout	Corresponds to a fixed time-out value in milliseconds, which is set at the beginning of the process. Once the time-out value expires, the extension drains all the events that are buffered within the reorder extension to outside. The time out has been implemented internally using a timer. The events buffered within the extension are released each time the timer ticks.	-1 (timeout is infinite)	LONG	Yes	No
max.k	The maximum threshold value for K parameter in the K-Slack algorithm	9,223,372,036,854,775,807 (The maximum Long value)	LONG	Yes	No
discard.flag	Indicates whether the out-of-order events which appear after the expiration of the K-slack window should get discarded or not	false	BOOL	Yes	No

Examples EXAMPLE 1

define stream inputStream (eventtt long, price long, volume long);
@info(name = 'query1')
from inputStream#reorder:kslack(eventtt, 1000L)
select eventtt, price, volume
insert into outputStream;

This query performs reordering based on the 'eventtt' attribute values. The timeout value is set to 1000 milliseconds

Akslack

reorder (Stream Processor)

This stream processor extension performs reordering of an out-of-order event stream.
 It implements the Alpha K-Slack based out-of-order handling algorithm (originally described in
http://dl.acm.org/citation.cfm?doid=2675743.2771828)

Syntax

akslack:reorder(<LONG> timestamp, <DOUBLE> correlation.field, <LONG> batch.size, <LONG> timer.timeout, <LONG> max.k, <BOOL> discard.flag, <DOUBLE> error.threshold, <DOUBLE> confidence.level)

QUERY PARAMETERS

Name	Description	Default Value	Possible Data Types	Optional	Dynamic
timestamp	Attribute used for used for ordering the events		LONG	No	No
correlation.field	Corresponds to the data field of which the accuracy directly gets affected by the adaptive operation of the K-Slack extension. This field is used by the Alpha K-Slack to calculate the runtime window coverage threshold which is an upper limit set for the unsuccessfully handled late arrivals		DOUBLE	No	No
batch.size	The parameter batch.size denotes the number of events that should be considered in the calculation of an alpha value. batch.size should be a value which should be greater than or equals to 15	10,000	LONG	Yes	No
timer.timeout	Corresponds to a fixed time-out value in milliseconds, which is set at the beginning of the process. Once the time-out value expires, the extension drains all the events that are buffered within the reorder extension to outside. The time out has been implemented internally using a timer. The events buffered within the extension are released each time the timer ticks.	-1 (timeout is infinite)	LONG	Yes	No
max.k	The maximum threshold value for K parameter in the Alpha K-Slack algorithm	9,223,372,036,854,775,807 (The maximum Long value)	LONG	Yes	No
discard.flag	Indicates whether the out-of-order events which appear after the expiration of the Alpha K-slack window should get discarded or not	false	BOOL	Yes	No
error.threshold	Error threshold to be applied in Alpha K-Slack algorithm. This parameter must be defined simultaneously with confidenceLevel	0.03 (3%)	DOUBLE	Yes	No
confidence.level	Confidence level to be applied in Alpha K-Slack algorithm. This parameter must be defined simultaneously with errorThreshold	0.95 (95%)	DOUBLE	Yes	No

Examples EXAMPLE 1

define stream inputStream (eventtt long,data double);
@info(name = 'query1')
from inputStream#reorder:akslack(eventtt, data, 20l)
select eventtt, data
insert into outputStream;

This query performs reordering based on the 'eventtt' attribute values. The batch size applied here is 20.