Essbase add views to existing aggregation

Question

Hello All,
I need help adding new views to an existing .csc file.
Here is the process:
1)     After cube data is loaded query tracking is enabled
2)     Users run their retrieves
3)     Aggregation is saved and new views are added to the existing views in the file
The MaxL I have for step 3 is the following:
execute aggregate selection on database application.database based on query_data dump to view_file myviewfile;
The following error displays with this script:
Viewfile [] already exists
Unexpected Essbase error 12141180
What do I need to do to add the views to the existing file?
Thank you

Answer 1

Thanks John - I wasn't sure of that myself and ran a quick test last night, your skepticism prompted me to re-run it.
The first time around I was sure it produced a file that contained all existing views plus the new ones (which isn't strictly the same as appending to the file, but produces the same results in this situation). However, when I retest in the cold light of day, it doesn't contain all the existing views - only some of them, presumably because they just happen to have been reselected by the new selection process.
There is some syntax in the Tech Ref which purports to ensure that existing views wind up in the new script, which may help the o/p:
execute aggregate selection on database app.db using views 0, 1, 2, 3, 4 with outline_id 12345 force display based on query_data force_dump to view_file filename;
The choice of 'display' or 'suppress' keywords determines whether the new file contains the existing views or not. There doesn't seem to be a shorthand for listing out the view ids, which is a bit of a pain. The view ids can be obtained from the original view file (they are the values on every other line from the third line onward) or as output from:
query database app.db list existing_views;
Never actually used this except just now to test it. It'd be a good idea to back up the existing view file regardless of approach. If I used the above syntax, I think I'd be inclined to use 'suppress' to produce a second script rather than modifying the original.