QlikView Task xml files for deleted tasks still exist

[mhouck01]

I'm trying to create a dashboard which shows the current tasks in QMC and who they are distributed to. The distribution lists for our tasks have changed frequently and they've been deleted and recreated a few different times.

I'm running into an issue because the C:\ProgramData\QlikTech\DistributionService\Tasks directory contains what seems to be ALL of the tasks that have ever been created. Even those tasks which were deleted and obsolete still have .xml files in this directory. So if a task with an old distribution list was deleted and then a new task was created, that document now has 2 tasks and I can't determine which is the correct one with the correct distribution list versus which is the old one.

Is there some way of telling which of the task.xml files currently exist in QMC versus those that have been deleted?

I've checked already to make sure that none of the tasks are orphans... I can't figure out how QMC can tell which are current tasks and which have been deleted.

[Peter_Cammaert · Accepted solution]

You're looking in the wrong place. The directory you're scanning at the moment is one of the many task-related directories that contains both current and historical run-time information. You'll even find files that contain momentary results (for example EDXResults: while you open a file, the QDS may already have written something else in it)

Task definition information is stored where the QMS can easily get access to it (QDS doesn't create task definitions, QMS does) It's called the Publisher Repository or QVPR. Although you can store a QVPR in an RDBMS, by default the information is stored in a set of xml files. Have a look in C:\ProgramData\QlikTech\ManagementService\QVPR.

Best,

Peter

[mhouck01]

Thanks very much. That was the piece I was missing.

[Anonymous]

Michael

Rob Wunderlich has already implemented a dashboard which I think gives exactly what you are looking for.

Take a look at QVPR Analysis

[mhouck01]

Very nice application there... I ended up doing much of the same for the app that I was working on.

The main purpose of my app was to look at access and usage of the various applications that we have deployed. So I pulled in session logs and then the task information and joined them together to answer questions like: Who no longer has access? Who has accessed which applications? Who has access but has never accessed the applications?

I also ended up linking out to our active directory server to pull in user and group information. It's a great help for all of our access administration and utilization analysis needs.

[Anonymous]

Getting who has access but hasn't used it by cross referencing with session logs. That's a really nice idea.

Hope you don't mind if I borrow it.

Thanks.