Contributor Synchronise Error: "...not a simple hierarchy"

[PhilE]

Hello all

I am unable to synchronise a particular Contributor model with Analyst as when I try to do so I get the following error:

"D-List 'xxxxx' has been selected as the E-List is not a simple hierarchy"

A couple of points about this:

This is not a new model, yet I've never had this error before

I have made some changes to the E-List in question but it is exactly the same as the E-List in another model, and that model synchronises just fine

Any help much appreciated

Regards,

Phil

[jan.herout]

You really should delete all items (but one) from the dlist that is used as elist in your model. The elist selected when creating the application is only PLACEHOLDER that is overwritten by the real elist later on so having more than one item in it has no purpose.

By doing so, you also effectively decrease size of the model that needs to be loaded into memory when syncing the application. In past I once saw that it was not possible to sync the application due to size of the model, which was caused by using the "real" elist structure as items in the placeholder.

[jan.herout]

Excerpt from Cognos official knowledge base>

Problem(Abstract)
Unable to Synchronize a 7.2 application.

Error Message -

elist has been selected as the E-List is not a simple hierarchy

Not all dimensions for this cube were packed successfully

elist has not been packed


Resolving the problem
This issue is caused by having a complex hierarchy in the elist within Analyst.

Ensure the elist in Analyst does not have a node present twice rolling up to two different levels.

[PhilE]

Excerpt from Cognos official knowledge base>

Problem(Abstract)
Unable to Synchronize a 7.2 application.

Error Message -

elist has been selected as the E-List is not a simple hierarchy

Not all dimensions for this cube were packed successfully

elist has not been packed


Resolving the problem
This issue is caused by having a complex hierarchy in the elist within Analyst.

Ensure the elist in Analyst does not have a node present twice rolling up to two different levels.

Hey, thanks for that! Looking at it with a clear head after the weekend, the E-List was as you stated, with a couple of items that I moved still existing in another roll-up.

[PhilE]

You really should delete all items (but one) from the dlist that is used as elist in your model. The elist selected when creating the application is only PLACEHOLDER that is overwritten by the real elist later on so having more than one item in it has no purpose.

By doing so, you also effectively decrease size of the model that needs to be loaded into memory when syncing the application. In past I once saw that it was not possible to sync the application due to size of the model, which was caused by using the "real" elist structure as items in the placeholder.

Cheers for the suggestion about deleting the analyst e-list.

You would then not be able to pull all of the data from Contributor for use in Analyst though, correct?

[jan.herout]

The trick here is to have at least two dlists. For example:

2_elist : {A}                                                                                -- this will be used as PLACEHOLDER
2_Profit_Center : {Profit center A = Profit Center B + Profit Center C}    -- this will be exported as ELIST int TXT and loaded as elist via CAC


You will use 2_elist as elist in application library, while 2_Profit_Center will be used in libraries for upload / download data from contributor.

Also, you should be aware that
- when runing dlink against Contributor, elist dimension will contain IDs of elist items, not their names!
- changing elist item ID is destructive => often you want to use fixed length string as elist item ID (eg. code of the profit center) while name of the item should be descriptive (eg. name of the profit center)

This is why I do not use dlist - export as elist option in Analyst very often and prefer to prepare elist manually.

[adityashah27]

just FYI... changing/modifying elist item id wont be destructive. i have done several times. if you delete and add new one is destructive.
you are right on using code or number in item id.
also curious to know why do you need to update elist manually, there are several ways to automate it end-to-end. share your process if want to brainstorm.

[jeffowentn]

I'm getting this error message about the simple hierarchy, yet the elist hierarchy is "simple" - there are a number of subtotals and one total, but each child only has one parent and every child is accounted for in a subtotal or total.  Does anyone have any other suggestions?

Also, I am interested to know what some people's processes are related to automation of the elist maintenance...we have a number of business departments that pull data back into Analyst from Contributor, etc...

Thank you, in advance, for your input, as we are in the midst of an upgrade to 8.4.1 with a tight deadline, and this sync issue in 7.3 is holding us up...

[ericlfg]

Hey Jeff,

Was there any work done on the e.list recently?  If so back track what was changed and make note.  I've seen this error once and it was corrected by fixing the e.list, as stated above in the TN.

[jeffowentn]

I'm not sure of any recent changes only b/c I don't "own" this application - trying to "fix" it so I can upgrade it...I'm going through, one level at a time, and adding back all the subtotals until I "break" it - we'll see where I get to, Eric.

[jeffowentn]

I ultimately just rebuilt all the calculations, one level at a time, and sync'ed after each calc to make sure it was still working ok...not sure what the issue was, but it is fixed, now...