TM1 Forum

Hi
I have task where all users must be able to see the base data and should create Sandboxes and play around and must not able to COMMIT.

Only one Main User must be able to Commit his sandbox to Base Data. After he commits ..all other Users will click reset to get fresh BASE Data

So,
How can i restrict Users from Commiting in TM1 9.5.1

Thanks for your time

I have no idea how to restrict it.

But maybe a simple idea solves your problem.

First, I understand all of your users are allowed to play around with the same slice of data (so it's not like UserA plays only with ProductA, UserB only with ProductB etc., but everybody can play with all products).

If so, maybe you should create as many copies of "base" as many users you have. Then you give every user WRITE access to her/his copy only and READ to the main BASE. You can also write a TI that refreshes all copies after your PowerUser changes something in the BASE.

I do not know if it will meet your needs (may depend on number of users), but still it is some kind of a workaround if restricting committing is not possible.

So... if i have 10 users and i write TI script identifying users and loading values to each on untill evey 1 is loaded with data. Will that degrate my performace or will take long time??
My data will be not very Huge... its presiced..

Thanks for the Advice...

I think this can be done using the Capability Assignment in the Server Explorer. Right click on the Server name in the Server explorer and you will find this option on the menu.

You can read more here:

http://publib.boulder.ibm.com/infocente ... D2001.html

Hi
Thanks for the link ..But after reading that i understand the If the Users given permissions to use Sandboxes and permission for creating their Own Sandbox then they will have Option for Commit.. .

But i dont want all the Users to Commit their Sandboxes to Base. Only Primary User must Commit to Base.
Is there any other ways to achieve this
My Old Idea is to grate a User Dimension... Each user have his own element in USer Dim and his data will be for him even he Commit..
But here it is a lengthy process and takes time for loading data for each user

Any other approaches you can advice me..

Thanks for ur time.

rkaif wrote:I think this can be done using the Capability Assignment in the Server Explorer. Right click on the Server name in the Server explorer and you will find this option on the menu.

You can read more here:

http://publib.boulder.ibm.com/infocente ... D2001.html

No. The relevant capability assignments determine whether users are set to use personal workspace versus direct write-back to base model and then whether users are able to create their own sandboxes. If you read what the OP is asking he is looking for a solution where users can do scenario analysis in sandboxes but are prohibited from committing the changes (i.e. writing back to the base model.)

vinnusea - the functionality you are looking for is basically not how the personal workspace and sandbox functionality was envisaged and intended to work. The primary use purpose of this feature was supposed to be for write-back budget contribution models. The permissions flow automatically FROM the base model TO the sandboxes. If a user has write access to the base model then they have write access in sandbox mode AND the ability to commit changes back to the base model. What you want to do if I understand correctly is for users to have only READ access to the base model but nevertheless have WRITE access to sandboxes. This is not possible. I agree with you though that it would be useful functionality in certain circumstances - the correct course of action to follow is to submit an enhancement request to IBM for this feature.

Here's an untested work-around that I think will work for you ...
1/ Set up your versions so that your single "write-back user" has WRITE access to the relevant version and all other users have only READ access
2/ Create a new version, for argument's sake call it "scenario analysis". Give all users WRITE access to scenario analysis
3/ Set up a TI to copy from the base version to scenario analysis and set this to run at regular intervals

I don't think in this case that it would be warranted to duplicate the data with lots of versions unique to each user as per jstrygner's suggestion. Probably a single additional version will do. You should still tell users not to commit back to the scenario analysis from their sandboxes but if they do then no harm would be done as it will be overwritten regularly from the base model in any case.

Make sense?

Yes lotsarm
your right .. You got my point ..
So creating a analysis Version will do and TI will update it in regular intervals. This could be a wise thing to do in my sinario..
will try to accomplish this and update .


Thanks for the advice and appreciated ur time for me..

Agreed with Lotsaram - that is why in my previous reply I said "I think".

Also if you want your users not to change the base version you may want to disable the Sandbox feature for all the users using the DisableSandboxing switch in your TM1S.CFG file.

Once again I am not saying you have to do it - but you may want to disable it when using multiple versions.

Yes Kaif
i think i need to disable sandbox for base version and need to create a new analysis version and give a Try...
Thanks for the reply

This will probably get listed as the "Stupid question from the new guy," but...

Depending on the resources needed for the server, would he be able to setup local servers for the users? It would be identical to the 'scenario analysis' server, except run on the user's local machine. This would give each user their own version to tool around in, and it could still update periodically from the 'base server'.

That is really only necessary if each user NEEDS their own version to mess around with. If having one central 'scenario server' works, that's probably the best solution.

aking wrote:This will probably get listed as the "Stupid question from the new guy," but...
Depending on the resources needed for the server, would he be able to setup local servers for the users? It would be identical to the 'scenario analysis' server, except run on the user's local machine. This would give each user their own version to tool around in, and it could still update periodically from the 'base server'.

Notwithstanding the technical difficlties of doing what you are talking about, unless you have full-blown Perspectives licenses for all those users you can't do that. The licensing model has changed drastically since IBM bought Cognos and the OP may not have the licenses to do that.

See? What'd I say, stupid question.

I Achived it...
Here is my Path

I have created a TI process which will Pull all The Groups in TM1 and Insert those elements Into a regular Dimension called USers . Each user have his own group
LIke Marry user will have Mary group..

my user Dim is created and In the Cube Securities

Mary will have write access to Mary Group and others None. So mary just see the mary cross section so Independent Base for mary in same way every user will have his/her own base.

And Loading data to each User is Down by Other TI process .. This will identify element by element in Users Dim and load data indivdually.

So all users play with their sandbox and even they commit it be be for his own BASE so not messing with Admin BASE.

My TI process for loading each USer is taking less time about 10-12 sec. so its Good..

And security is Dynamic its works like

"IF USer name matches with Group Name ,Then Write Else None"

Its done now ..

Thank you all for ur time.

Hello

I face the same requirement that a COMMIT should not be possible in a sandbox. Or possible to be set up for each security group (a new capability).

This is a problem in the TM1 application at hand for 2 particular users.

Unless someone of you knows a better solution, I will most probably end up copying the data to a new element in the Years dimension (for 3 cubes, the other cubes are rules-driven).
In fact, I will have 2 dummy years, 1 for user A and one for user B. These users will use the 3 cubes and change parameters and percentages for a given year, as a kind of "What if analysis".

It's not ideal though.

Wim

Hi vinnusea, you can also use Approval Hierarchy to control your users' access to the data (if you use TM1 Contributor).

Using a very simple Approval Hierarchy, each user/group is assigned to a separate node in the approval hierarchy. Let them open the node and play with data.

In the end, copy the "base" node to overwrite all other nodes (using a TI process).