If you are unable to create a new account, please email support@bspsoftware.com

 

News:

MetaManager - Administrative Tools for IBM Cognos
Pricing starting at $2,100
Download Now    Learn More

Main Menu

Access issue using access tables?

Started by Preethi, 09 Mar 2010 11:04:01 AM

Previous topic - Next topic

Preethi

Hi All,

We have 4 cubes in an application, categorized as

Company 1 - Cube 1, ( Dim are 1 Cost cal, 2 Cost Centers_1, 2 Serial_1, 5 Versions)
Company 2 - Cube 2, (Dim are 1 Cost cal, 2 Cost Centers_2, 2 Serial_2, 5 Versions)
Company 3 - Cube 3, (Dim are 1 Cost Cal, 2 Cost Centers_3, 2 Serial_3, 5 Versions)
Company 4 - Cube 4, (Dim are 1 Cost Cal, 2 Cost Centers_4, 2 Serial_4, 5 Versions)

I cannot have cost centers as an elist, because the user belonging to company 1 wants to load the data for all cost centers_1 @ web level using Local link....for each company, we have not more than 2 users.

User 2 Should see only  Cube 2

User 1 and User 3 cubes 1, 3,

User 4 and User 1 should see Cube 4

Reviewer should see all cubes

I used company as an elist, and controlled the access to cubes thru access tables.

Issue is: When user 1 clicks Company 2 node, he can see Cube 4, and Cube 3 cube also...

My Req is When the User 1 clicks company 2 node, he should see only company 2 ....


Please let me know how to implement this and also let me know is there a better way to do?

Thanks,
Preethi  :)


                                           




Milt

Hi, if I understand the scenario correctly then you can only achieve this by splitting the cubes into seperate contributor applications. Then you can set access to each application via the 'rights'.

It might be possible to construct an access table that will provide a solution. But, this depends on how your data is organised

>Milt

craig_karr

Provided that there are a limited number of cost centers and that the model is fairly small, you could perhaps combine company and cost center to a hierarchy e.list and enable the multi e.list items options. Then the users that have access to a company can open all the cost centers in that company at the same time and that should also make it possible to execute a local link to all cost centers in one company at once

branko

A nice topic :-) just
4 cubes for the whole solution....
1 cube for a company...
2 users for one company .. .

Well, users always have a "strange" & "complicated" reqs. sometimes no one knowing the reasons behind :-).. I would try & started this way :

1.st - a clarification meeting, discussion  - to find out :

- the reasons why they want such rights / maybe to discuss the possibilites, suggest better approaches, based on the discovered reasons behind...

- if they really need a contributor solution for that :-) i would used it if there where - more users ? much inputs needed to be collected, reviewed - what about the analyst solution for them (maybe they already have & use analyst) the user for company 1 - that would like to load data for all cost centers - he will load the whole model alone - would not he be happier with the analyst ? There are more possibilities - a solution can be a combination of Analyst & XLS templates (to collect data) or a Contributor and Analyst or a Contributor and XLS templates (not everyone necesarrily have to have a contributor client - and in some cases alternatives suites for some users better ..)


2. nd - surely not to have everything in one contributor model (even if the model is small) but because of the criss-cross rights :-). If I were "forced to do such kind of solution" I would separate models (based on security, rights ... of course the logic too :-)).. but trying to have not many unclear acc. tabs.

3. rd - a rework of that solution - like mentioned by craig_karr could help too - but it is about meeting & discoussions - if they are the users who can make a compromises - then look back to 1st step - if not - then change the model that way could also challange a wave of criticism (from the users side..)

Preethi

Thanks, for sharing ideas.
Surely, it will help me.

The problem we have is analyst licsense and the users dont want to load the data in analyst cubes becoz of payroll info... they like to do it in Contributor without Cognos administrator help, using local links.

This works fine when the cube size is less than 2.5 million cells, if its above than that ...user is unable to load the data in to cubes.

Thanks,
Preethi

branko

Well,
reg. the problem with analyst license  - (if the problem is price?...) - not everyone, of course need to have Analyst licence it could be a combination of what i mentioned e.g. 1 Analyst & X users based on XLS templatet (maybe happy - they could still use the XLS :-)) and the one with analyst would consolidate the input in a minute...

reg. users do not want to load the data in analyst because of payroll info ? (what payroll info ?) they just simply like contributor ? - you can make analyst solution for them much more attractive usign CP Manager or Add-in for Excel ? what ever they like...(so... i did not mean to force them to know the exact d-cube names, open them in analyst manually & enter data & run a link manuallly or a macro manually :-) for that i would use Manager or Excel - whatever they like more...

However, the main things i do think are relevant :

- it does not have to be 1 contributor model (you have a lots of possibilities offer them other tools & combinations of tools or create several models)

- try to discuss with them (the reasons why they intend on having only 1 contribu. model, criss-cross rights, want to use just contributor...)

BTW. using analyst the cube size could be much more than 2,5 million cells to load the data (even ten times more...) - another limitation for contributor ...(which is primarily a tool for data collecting & not importing & processing the data). What would they do next - after importing the data do contributor ? (print them ? ... download back to analyst ? :-).... publish or ?)

Have a nice weekend...





Preethi

#6
Hey Branko,

Thanks a lot....

After loading the data through excel files, they submit the data for review.

Once review is done we will publish the data for BI team to create reports.

Have a great weekend :-)

Regards,
Preethi