Is the session being run via a single call to fireUntilHalt()? -W On 18/09/2013, Alexander Wolf <mail(a)alexander-wolf.net&gt; wrote: > [Drools Version 5.5.0 Final] > > Hey - > > I got an event E1 that is only valid, if the latest event E2 following E1 > within 2 minutes has the value 1 > --> I want have the rule fire 2 minutes after E1 but only if E1 is valid > > This is my rule: > > rule "inform about E1" > when > //event (T1) is the initial trigger > $event1 : Event(type == EventType.T1) > //there is an event (T2) with value 0 between 0,2m after doorClosed > $event2: Event(type == EventType.T2, value == 1, this after [0, 2m] > $event1, $timestamp : timestamp) > //there is no newer event (T2) within the timeframe > not Measurement(type == EventType.T2, this after [0, 2m] $event1, timestamp >> $timestamp) > then > //print info > log(drools, "E1 valid"); > end > > An example of Events: > > 12:00:00 - E1 > 12:01:00 - E2 ( value = 0 ) > 12:01:10 - E2 ( value = 0 ) > 12:01:40 - E2 ( value = 0 ) > 12:01:50 - E2 ( value = 1 ) > 12:02:10 - E2 ( value = 0 ) > > I would expect the output: [log() does log the clock-time when my rule > fires) > > 12:02:00 E1 valid > > But what I get is: > > 12:03:50 E1 valid > > > So I see that the late coming E2 (@12:02:10) is correctly ignored, the rule > result is basically correct, but the rule is fired much to late. (actually > 2 minutes after $event2 and not as expected 2 minutes after $event1). > > Could someone give me a hint what I did wrong? > > Regards, > Alex > _______________________________________________ rules-users mailing list rules-users(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/rules-users

Thank you Davide! For now, as a work around I split this into two rules, 1. rule : register the initial and second event and after (timer:2m) fire an internal event 2. rule : register internal event and then do the look behind (last 2 minutes) --> assert that the original event was valid --> this is not nice code and might cause further problems, so I am really looking forward to get a bug fix ;) @W: I call "fireAllRules" every time I insert a new event. (is this bad? This seemed to works better (in terms of predictability (unit tests...) of outcome) than all my attempts to use "fireUntilHalt", w and w/o new Thread ...) especially with rules that use cron timers. On 19.09.2013, at 10:23, Davide Sottara <dsotty(a)gmail.com&gt; wrote: > It is a bug. > The 2s delay is computed from the moment the rule is activated (by E2 at > 12:01:50). > rather than taking into account the correct "zero" timestamp ($event1). > Thanks for reporting this, I'll open a JIRA > Davide > > On 09/18/2013 12:29 PM, Wolfgang Laun wrote: >> Is the session being run via a single call to fireUntilHalt()? >> -W >> >> On 18/09/2013, Alexander Wolf <mail(a)alexander-wolf.net&gt; wrote: >>> [Drools Version 5.5.0 Final] >>> >>> Hey - >>> >>> I got an event E1 that is only valid, if the latest event E2 following >>> E1 >>> within 2 minutes has the value 1 >>> --> I want have the rule fire 2 minutes after E1 but only if E1 is >>> valid >>> >>> This is my rule: >>> >>> rule "inform about E1" >>> when >>> //event (T1) is the initial trigger >>> $event1 : Event(type == EventType.T1) >>> //there is an event (T2) with value 0 between 0,2m after doorClosed >>> $event2: Event(type == EventType.T2, value == 1, this after [0, 2m] >>> $event1, $timestamp : timestamp) >>> //there is no newer event (T2) within the timeframe >>> not Measurement(type == EventType.T2, this after [0, 2m] $event1, >>> timestamp >>>> $timestamp) >>> then >>> //print info >>> log(drools, "E1 valid"); >>> end >>> >>> An example of Events: >>> >>> 12:00:00 - E1 >>> 12:01:00 - E2 ( value = 0 ) >>> 12:01:10 - E2 ( value = 0 ) >>> 12:01:40 - E2 ( value = 0 ) >>> 12:01:50 - E2 ( value = 1 ) >>> 12:02:10 - E2 ( value = 0 ) >>> >>> I would expect the output: [log() does log the clock-time when my >>> rule >>> fires) >>> >>> 12:02:00 E1 valid >>> >>> But what I get is: >>> >>> 12:03:50 E1 valid >>> >>> >>> So I see that the late coming E2 (@12:02:10) is correctly ignored, the >>> rule >>> result is basically correct, but the rule is fired much to late. >>> (actually >>> 2 minutes after $event2 and not as expected 2 minutes after $event1). >>> >>> Could someone give me a hint what I did wrong? >>> >>> Regards, >>> Alex >>> >> _______________________________________________ >> rules-users mailing list >> rules-users(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/rules-users >> > > _______________________________________________ > rules-users mailing list > rules-users(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/rules-users