Performance problems with consolidated elements

    Performance problems with consolidated elements

    I have a speed problem updating views.



    I have a cube with 10 dimensions. These are Plan, Version, Date, Datatype, Officers, Cost Center, Branches, Areas, Personal Measurement and Concepts.



    Plan, Version, Date, Datatype and Cost Center have only one element each.



    Officers have 174 elements. An item called "To every element" 172 number format elements (1, 2, 3 ...) and one called "Total officials” (which consolidates previous 172 elements.

    Branch has 25 members and three consolidated. One of consolidated ("Total branches ") is the father of the other two bound (" Metropolitan " and "Interior"). And these two are the parents of the base elements.

    Areas have 16 elements one consolidated and 15 base elements. Consolidated is called "Total area" and includes 14 of the base elements.

    In Personal Measure I have several elements. Among them I have a structure for the payment of salaries, which I explain below.

    Loss on remuneration © (Father of Monthly Salary Cost and Social Provisions)
    © Monthly Salary Cost (Father of the base elements below)
    Salary no social security taxes ( B)
    Social charges (B )
    © Social Provisions (Father of the base elements below)
    A Provision (B )
    Vacation Pay Provision (B )
    CJPB Provision (B )



    In Concepts I have 5 items and 2 consolidated basis. The structure is as follows:
    Total © (Father of Total Fixed and Variable)
    Total Fixed © (Father of Base, Presenteeism and Tickets)
    Base ( B )
    Presenteeism (B )
    Tickets ( B)
    Variable ( B)
    For every concept (B )



    If I want to pull a view for a base element of the first dimensions (that have one dimension), and the maximum consolidation for the rest of the dimensions, PALO is processing the information indefinitely. If I change one of the consolidated elements by a base element, brings the right information.



    I have 42 rules and 31 are in use in the view.



    I already try using PALO.MARKER.

    In some moment I have one additional dimension with 5 base elements. When eliminated this dimension the system start to work so much better.

    Do you have some solution for my problem? Do you think I have to work in the rules? Or the problem is the size of the cube?
    Hi clau,

    I'd guess it is problem with rules. Is it possible that some of them do not have markers defined? Usually when you define marker in one rule you must have to define them in all other in the same cube and source cube of these rules.
    You didn't write if you are using rules on base cells. (B: or N: rules)
    Other common problem is that if you define rule on one target element that has parent element in the dimension and you use that parent element in your view. Then the rule is calculated as B: rule no matter it is general rule.

    Size of the cube and dimensions is usually not problem

    regards

    Jiri
    Thanks for your answer.

    I am using markers only in some rules. My problem is that I don`t know how to use the markers in all the rules.
    I have the idea that I can only use the PALO.MARKER if I have defined the same dimensions in target and in source. My problem is that is not the structure of my rules.

    For example, I have my salaries charged in a specific data ("Start") and I wana see this salaries in all months (oct-13, oct-14, etc), with the corresponding actualizations (increse %). In that case I have this rule:

    ['Salaries','Financial','All cost center'] = N:['Salaries','Financial', 'Start','All cost center'] * ['Month'] * (1 + ['Actualization','All officers','All areas','All branches','All concepts'])

    In this case, I don`t have the same dimensions defined in the target and in the source.

    I am going to explain only with a part (for more clarity).

    ['Salaries','Financial','All cost center'] = N:['Salaries','Financial', 'Start','All cost center']
    In this case, the dimension Data is free in the target and especific in the source.

    Other question... Can be a problem with the size of the view?

    Regards,

    Claudia.
    yes. it is problem of markers that they cannot be defined for all types of rules.

    several notes to your rule:
    ['Salaries','Financial','All cost center'] = N:['Salaries','Financial', 'Start','All cost center'] * ['Month'] * (1 + ['Actualization','All officers','All areas','All branches','All concepts'])


    - would be better in this forum to write rule targets and sources in format ['dimension':'element'] so I could understand it better without knowing complete dimension
    - what is planned type of element 'All cost center'? It is is consolidation then the rule will not work because the rule is restricted to base cells only ( N: )
    - what dimension comes the 'Month' element from? Date?
    - element 'Start' you used in source is from dimension Date or Data? In your first post you wrote that dimension Data has only one element
    - are you sure that source ['Actualization','All officers','All areas','All branches','All concepts'] is correct? It seems to be referencing consolidate cell sum of all 'Actualizations' over 'All officers' etc.- I'd expect you want to multiply with something specific for each source value ['Salaries','Financial', 'Start','All cost center']

    In your first post you wrote:
    In some moment I have one additional dimension with 5 base elements. When eliminated this dimension the system start to work so much better.

    I understood this that when you added some elements into one of dimensions, calculation was much longer. What dimension was that?

    Jiri