You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@purge(enable='true', interval='10 sec', idle.period='1 min')
partition with ( lineId of RaceWaypointStream )
begin
from RaceWaypointStream
select lineId, str:concat(waypointId, '$', convert(timestamp,'string'), '$', vehicleId,'$',lineId) as ranking, vehicleId, timestamp
insert into RankStream;
from RankStream#window.unique:ever(vehicleId)
-- select timestamp, lineId, convert(list:sort(list:collect(ranking), 'DESC'), 'string') as sortedRank
select timestamp, lineId, ranking as sortedRank
insert into RankOutStream;
end;
a producer with 100 kafka-producers running each in a thread produce 1 events per 10 ms. all events have different content.
as output we first see
that the loggers log in the expected order:
IN, MID, OUT, IN, MID, OUT, ....
at some point we see:
OUT, OUT, OUT, ... with the same event being duplicated up to 50-80x which is causing a congestion of the test system.
Question: is this an issue with the script ? or a bug ?
Related Issues:
The text was updated successfully, but these errors were encountered:
Description:
unique:ever window emits unexpectedly many duplicats
Affected Siddhi Version:
siddhi-tooling-5.1.0
OS, DB, other environment details and versions:
win 10
Steps to reproduce:
@app:name("RaceRank")
@app:description("Description of the plan")
@sink(type='log', prefix='IN')
@source(
type='kafka',
topic.list='kafka_waypoint_topic',
partition.no.list='0',
threading.option='single.thread',
group.id='group',
is.binary.message='true',
bootstrap.servers='localhost:9092',
@Map(type = 'json'))
define stream RaceWaypointStream (timestamp long, lineId string, waypointId string, vehicleId string);
@sink(type='log', prefix='MID')
define stream RankStream(lineId string, ranking string, vehicleId string, timestamp long);
@sink(type='log', prefix='OUT')
-- @sink(type='kafka',
-- topic='kafka_rank_topic',
-- is.binary.message='true',
-- bootstrap.servers='localhost:9092',
-- is.synchronous='true',
-- @Map(type = 'json')
-- )
define stream RankOutStream(timestamp long, lineId string, sortedRank string);
@purge(enable='true', interval='10 sec', idle.period='1 min')
partition with ( lineId of RaceWaypointStream )
begin
from RaceWaypointStream
select lineId, str:concat(waypointId, '$', convert(timestamp,'string'), '$', vehicleId,'$',lineId) as ranking, vehicleId, timestamp
insert into RankStream;
-- select timestamp, lineId, convert(list:sort(list:collect(ranking), 'DESC'), 'string') as sortedRank
select timestamp, lineId, ranking as sortedRank
insert into RankOutStream;
end;
a producer with 100 kafka-producers running each in a thread produce 1 events per 10 ms. all events have different content.
as output we first see
that the loggers log in the expected order:
IN, MID, OUT, IN, MID, OUT, ....
at some point we see:
OUT, OUT, OUT, ... with the same event being duplicated up to 50-80x which is causing a congestion of the test system.
Question: is this an issue with the script ? or a bug ?
Related Issues:
The text was updated successfully, but these errors were encountered: