Saved report layouts reverting back to default view

[Former Member]

Hi,

I would like to know what can cause a saved (pre-existing) A-OLAP report to lose its customized layout and revert back to the default view of its underlying BEx query.  Our BI A-OLAP uses BEx queries as data-sources, connecting to them via BICS. Platform version: BI 4.0.5.1

Every so often we receive user complaints that they have lost the customized layout of their saved reports.

1. What Bex query property settings, in query designer and in BW, can affect pre-existing OLAP Analysis Workspaces?

2. What changes (within an A-OLAP workspace or in BI, Bex, BW, CMC etc.,) can make an AOLAP report layout revert back to default view?

Thanks

[I_MCA]

Hello j n

What you are describing is impossible! An AOLAP workspace cannot "lose" it's layout. When you first create an AOLAP workspace it uses the default layout as defined in the BEx Query... but from that point on the workspace is largely decoupled from the Query. Think of the AOLAP workspace as being like a Query View. AOLAP has no concept of reverting to a default layout.... and yet it appears to be happening.

With that in mind there are several things to consider...

1) Are the AOLAP workspaces being opened/saved using the BI Launchpad or via a portal with iViews?

2) When the AOLAP workspace is saved, is it being saved where the user expects it to be saved (ie if they go to the Documents tab in the BI Launchpad can they navigate to where they saved their workspace, open it and confirm that it was indeed saved)?

3) Is there any backup/restore taking place where content is being restored with older versions?

4) When a workspace resets to the default layout, what is the date/time stamp on the workspace? Does it correlate to when the user last saved the workspace?

5) Is it ever possible that a user has rights to save a workspace to a particular folder at some times but not other times?

6) Are the complaints typically from the same users and on the same dates? Are there other users with similar work patterns who have never had a problem?

I hope these provide some areas of exploration.

Kind Regards

[Former Member]

Thanks Ian. I've read some of your older posts on BOBJ and was hoping to hear back specifically from you.

Out of your listed areas of considerations #2 may warrant some investigation. We will take it to BASIS.

But I still would like to know *exactly* what changes to the Bex query affect saved workspaces and what changes do not.

In one of your responses to an older discussion thread, you had stated "...when you open a saved analysis workspace that was created by Analysis OLAP, we connect back to the query for the core data definition but we ignore everything else...".

What part of the Bex query constitutes this core data definition?

[I_MCA]

Hi j n

There are no changes that can be made to the BEx Query that would cause AOLAP to change its layout or any customized work that was done by a user. Instead there will be warning messages. For example if a member is removed from the query which was part of a user defined calculation... or if a hierarchy was replaced by a different hierarchy in the query... these kinds of changes will cause AOLAP to issue warnings because it is not able to "redraw" the customized layout that it was last saved with. It will never revert to a default layout... worst case is that the workspace cannot be opened and whatever change was made to the query would have to be reverted.

Other query changes will not impact existing AOLAP workspaces, eg changing the format of numbers, conditional formatting... these can be done in AOLAP itself so these query changes would only impact new AOLAP workspaces and not existing ones.

It would be useful if you can identify any pattern to these reports about the workspaces getting reset. It would also be useful to confirm if the AOLAP workspace really has "reset" to the default layout defined in the query designer (ie if you create a new AOLAP workspace against the query is it really the same as the "reset" workspace?)

Of course, it is certainly possible that you have identified an issue that we have not seen before... but the folks I spoke to here had the same opinion that they could not understand how AOLAP could "reset" itself.

Let us know if you discover any common denominators to the reports.

Regards