VUE files in the TM1 data directory
Posted: Sat Feb 25, 2017 7:43 pm
Hello all,
If you create a public view on a cube, the corresponding vue file will be placed in the TM1 data directory in a folder specific to the cube.
However, when a TI process creates a cube, subsets and a view, the vue file will be put directly in the data directory, not in the cube folder.
When the TM1 model is restarted, however, the cube folder is created and a vue file is put in that folder. But the original vue file straight in the data directory, is not removed.
I find this all annoying. Not only should TM1 create the folder and put the vue file in it, even if it cannot, it should remove the original vue file to clean up objects.
I observe it with the latest version of TM1 but probably it's been there for a long time.
What do you think ?
Obviously I don't want to run a startup chore to delete *.VUE files.
TM1 should not behave like this in the first place.
We want to prevent this thing from happening rather than clean up the consequences.
TI code to replicate the problem - FILL IN vDim1 and vDim2 with existing dimension names:
If you create a public view on a cube, the corresponding vue file will be placed in the TM1 data directory in a folder specific to the cube.
However, when a TI process creates a cube, subsets and a view, the vue file will be put directly in the data directory, not in the cube folder.
When the TM1 model is restarted, however, the cube folder is created and a vue file is put in that folder. But the original vue file straight in the data directory, is not removed.
I find this all annoying. Not only should TM1 create the folder and put the vue file in it, even if it cannot, it should remove the original vue file to clean up objects.
I observe it with the latest version of TM1 but probably it's been there for a long time.
What do you think ?
Obviously I don't want to run a startup chore to delete *.VUE files.
TM1 should not behave like this in the first place.
We want to prevent this thing from happening rather than clean up the consequences.
TI code to replicate the problem - FILL IN vDim1 and vDim2 with existing dimension names:
Code: Select all
vCube = 'test';
vView = 'Wim';
vSubset = 'Wim';
vDim1 = '';
vDim2 = '';
CubeCreate( vCube, vDim1, vDim2 );
SubsetDestroy( vDim1, vSubset );
SubsetDestroy( vDim2, vSubset );
SubsetCreateByMDX( vSubset, '{TM1FilterByLevel( TM1SubsetAll( [' |vDim1 | '] ), 0)}');
SubsetCreateByMDX( vSubset, '{TM1FilterByLevel( TM1SubsetAll( [' |vDim2 | '] ), 0)}');
ViewCreate( vCube, vView );
ViewSubsetAssign( vCube, vView, vDim1, vSubset );
ViewSubsetAssign( vCube, vView, vDim2, vSubset );
) does not lose time when using those extra cubes.
