version lock

[ravi]

Hi,

I have a question on version lock.

Is it Ok to lock version prior to transfer of data from that version into another one?
Will it cause any issue with successful transfer?

Please advise.
Ravi

[lotsaram]

Is it Ok to lock version prior to transfer of data from that version into another one?
Will it cause any issue with successful transfer?

Yes it will cause an issue, namely the element will not be writable. If an element is locked then no updates are possible even by admin users or by TI processes to any intersection of the locked element in all cubes. Element locking is an extra layer of security on top of the data security model, it can be a very useful addition for requirements where even admins must not be able to change actual or a past forecast version should not be able to be updated by any user.

The major usability issue with element locking is that there is no programmatic interface to lock/unlock and this must be done manually via the subset editor.

[ravi]

Hi Lotsaram,

Thanks for reply.

I did a testing :

1) I created a dummy version as FY SeptOL
2) Transferred the forecast data from FY JulOL(Source) to FY SeptOL(target)
3) FY SeptOL(target) Security-> lock
4) data transfer was successful- without any causing any issue.

Anything is wrong in the above steps?
please advise
Regards,
Ravi

[David Usherwood]

Ravi did say he wanted to transfer _from_ the locked version - which he can.
I agree with you about the limitations of locking - but you didn't mention that the UI doesn't tell you what element, in what dimension, is locked! ('No write access to element '*' - ). Because of that I prefer to use normal element security but as you say that doesn't stop admins making silly mistakes - but surely they don't

[lotsaram]

Ravi did say he wanted to transfer _from_ the locked version - which he can.
I agree with you about the limitations of locking - but you didn't mention that the UI doesn't tell you what element, in what dimension, is locked! ('No write access to element '*' - ). Because of that I prefer to use normal element security but as you say that doesn't stop admins making silly mistakes - but surely they don't

I read Ravi's question as wanting to transfer data TO the locked version. On a re-read I see that he wants to copy FROM a locked version is which case obviously there is no issue.

In terms of finding locked elements I disagree. It is pretty simple to either zero suppress the }ElementProperties cube or use this piece of MDX to create a subset. (Assuming of course that you have a fairly good idea which dimension will contain any locked elements, provided only admins have the ability to lock elements then this should be a no brainer. I have seen plenty of situations where the business either doesn't fully trust the admins (or themselves) or where IT won't support a system unless there is a mechanism to prevent changes of say actual or budget even by administrators.

Code: Select all

{FILTER(
  {TM1SUBSETALL( [<dimension>] )},
  [}ElementProperties_<dimension>].([}ElementProperties].[LOCK]) <> ""
)}

Of course it would be nice to be told WHY a cell is not updateable with a reference to which security object and maybe which rule line, etc. but I don't think this is unique to element locking as a cell not being writable could come from cube security, element security, cell security, cube rule, non-leaf intersection, element locking, ... and you never get told where the issue is (with the exception maybe of" element type is consolidated" or "rule is in effect") tracing this is up to the developer and knowing the model.

Hi Lotsaram,

Thanks for reply.

I did a testing :

1) I created a dummy version as FY SeptOL
2) Transferred the forecast data from FY JulOL(Source) to FY SeptOL(target)
3) FY SeptOL(target) Security-> lock
4) data transfer was successful- without any causing any issue.

Anything is wrong in the above steps?
please advise
Regards,
Ravi

No, se above I just misread your post and thought you wanted to update a locked version not read from it.

[ravi]

Thank you- Lotsaram.

Regards,
Ravi

[David Usherwood]

@Lotsa, I must admit I had forgotten about the LOCK attribute and/or had thought it wasn't updated properly. I will check it out. I still think that the message about not being able to write to element * is a particularly unhelpful one.