Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 
Not applicable

Triggers on multiple events completed has stopped working

A couple of days ago a trigger for a task setted on multiple events completed has stopped working without any change on task schedules. Looking in the logs  I found that file "Root_YYYYMMDD.txt" has the following entries:

4/5/2012 02:42:06.4842657    Information    TaskCompletedTrigger 03a856ca-06fc-4afe-8098-80ad61a7901b <1> of <4> fired. AndTrigger: 83c9232b-e1f6-4b85-a68a-ff40084a7d20

4/5/2012 02:42:06.4842657    Information    TaskCompletedTrigger c3b0d8ba-f3ee-4fc8-a1d1-f163a14cdfc5 <2> of <4> fired. AndTrigger: 83c9232b-e1f6-4b85-a68a-ff40084a7d20

4/5/2012 02:42:06.4842657    Information    TaskCompletedTrigger f263eb79-83f7-4fff-abcc-63771fa7bf66 <3> of <4> fired. AndTrigger: 83c9232b-e1f6-4b85-a68a-ff40084a7d20

4/5/2012 02:42:06.4842657    Information    TaskCompletedTrigger d15a0334-e2aa-4b1b-9c48-e879f8456492 <4> of <4> fired. AndTrigger: 83c9232b-e1f6-4b85-a68a-ff40084a7d20

4/5/2012 02:42:06.4842657    Information    Time constraint encountered for AndTrigger: 83c9232b-e1f6-4b85-a68a-ff40084a7d20. TaskCompletedTrigger kept with ID c3b0d8ba-f3ee-4fc8-a1d1-f163a14cdfc5 and new StartTriggerTime (for AndTrigger): 4/5/2012 2:42:06 AM

Looking the xml configuration files, I found that the trigger id 83c9232b-e1f6-4b85-a68a-ff40084a7d20 is the trigger that stopped working.

The trigger's time constraint is set on 360 minutes. All the previous tasks ended in less time. I guess that a previous run of the trigger was alive and waiting.

After the first fail, it fails consistently every day. When something similar happened in the past, we fixed it re-creating the trigger. These

Any ideas?

14 Replies
garystrader
Partner - Creator III
Partner - Creator III

This has happened to me.  Same situation - multiple event trigger, long trigger time constraint, and tasks fully reloaded every 15 minutes, well below the time constraint.  It would work every 15 minutes for a day or two, then at some random time it would fail to fire the trigger.  We provided information to QlikTech Support but they were never able to reproduce.  Have you filed a support case?

I should add that I was using Server V10 SR2 and SR3.  What version are you using?

Not applicable
Author

Gary, we're using  QV10SR3. Is good to know that we aren't alone. We filed a support case, but never had an answer. Next week we will be collecting more information with our local Qlikview partner in order to fill a new case.

Not applicable
Author

I found a trigger file in C:/ProgramData/QlikTech/DistributionService/Triggers with status "Unhandled" for each trigger that doesn't work. Looking the date on the files, I guess that this file corresponds with a previous run of the dependent task (may be a manual run) and QV isn't handling well the trigger's expiration.

I tried deleting one of the files and the following time the trigger works fine.

So, I made a Python program that checks Triggers_*.xml files and move to another directory the triggers files with problems (status unhandled or trigger definition doesn't exists in current configuration).

You can get the program here http://goo.gl/gaM0B. You need Python 2.7 in order to run the program.

Usage: qvandtriggerreset.py [options]

Options:

  -h, --help            show this help message and exit

  -t TRIGGER_DIR, --triggerdir=TRIGGER_DIR

                        Directory where QV stores trigger files (must exists).

  -c CONFIG_DIR, --configdir=CONFIG_DIR

                        Directory where QV stores task and trigger

                        configuration files (must exists).

  -a ARCHIVE_DIR, --archivedir=ARCHIVE_DIR

                        Destination directory for trigger files to be reseted.

  -p TN_PATTERN, --tasknamepattern=TN_PATTERN

                        Task name pattern to lookup for triggers.

  -l LOG_FILE, --logfile=LOG_FILE

                        Log file (stdout if omited).

The idea is to run this program before a batch process begins. You can specify the tasks to include in the search using a regular expression (TN_PATTERN). For example, if tasks are named PROD ... (for all production tasks), you can reset all triggers with problems for that environment using -p "PROD.+" before chain of tasks starts.

Example (for ALL tasks):

python.exe qvandtriggerreset.py -t "C:/ProgramData/QlikTech/DistributionService/Triggers" -c "C:/workspaces/qvandtriggerreset/data/QVPR" -a "C:/ProgramData/QlikTech/DistributionService/Triggers/archive" -p ".+"

This is not a solution, but a workaround. I hope it is usefull until a real solution.

garystrader
Partner - Creator III
Partner - Creator III

This is really great.  You've inspired me to download python to try this out.

Your troubleshooting work should help QlikTech Support narrow down on what the real problem is and hopefully fix it.  Maybe they've already fixed it in V11.

danielrozental
Master II
Master II

Gary, do you have a case number that I can reference with QT support?

garystrader
Partner - Creator III
Partner - Creator III

The case that I previously worked?  I don't have access to it anymore because I've changed companies.  You could give them my name, they know me pretty well.

s_uhlig
Partner - Creator
Partner - Creator

Hi,

do you have any news on this matter?

Seems not been fixed with SR4.

Not applicable
Author

Not yet.

Not applicable
Author

we also are using Q10SR4, and have similiar issues with the Triggers on Multiple events