Saturday, September 24, 2016

Thoughts/Info on Essbase/EssCS after Oracle Open World 2016

I was at Oracle Open World last week and have some notes to share on the upcoming EssCS (Essbase Cloud Service) product and on Essbase in general.

EssCS will be Infrastructure as a Service (IaaS) product and will be priced on the number of cores and memory.  I believe there will be a metered and an unmetered pricing as well (metered meaning pay per cycle or some other usage measure).  According to presentations at Open World, which did have safe harbor statements meaning "do not make decisions based on this info as it could change", there will be options for 1, 2, 4, or 16 physical cores ("OCPU's") and 7.5 to 15 Gb of RAM.  In addition, it will be an updated version of Essbase that is not the current on-prem version.  It will feature, among other things:


  • The new Java Agent running in Weblogic that moves security from the essbase.sec file to a relational database.
  • Simplified security
  • ASO / BSO / Hybrid
  • Sandboxing and scenario management - what if's without copying data with workflow for approval to merge into given scenarios
  • Cloud-based UI
  • Ability to specify/create an Essbase cube from Excel 
  • A cube template library ("app store" type of thing)
  • A web-based outline editor (though most editing should be done in Excel)
  • EssCLI (Essbase Command Line Interface) - a sort of EPMAutomate for Essbase
  • The Essbase Java API and a new REST API (which is currently being engineered)

I do not remember hearing any dollar amount for EssCS at Open World.  I expect availability in the next 3 to 6 months though it wouldn't surprise me if it were to slip further.

As far as on-prem Essbase updates, I would expect that the updates we see in EssCS will go on-prem as part of the EPM 2017 on-prem release which Oracle currently believes will be delivered late in 2017 (also subject to safe harbor, etc).

As far as how Oracle is selling Essbase, Essbase is now firmly in the BI organization and is being sold by the BI reps; EPM reps do not generally sell Essbase.  To the Essbase team, EPM is important as they are an internal customer, but EPM is not their only customer.  As such, I saw at least one presentation that promoted the idea of customers writing custom planning applications using Essbase.  While some people I talked with thought that approach muddled the EPM message for customers, I see it as a situation where if they don't compete in the custom market, then someone else will.  As someone who frequently is involved in complex applications where the EPM Planning product may not be a fit, I am thrilled to see that message from Oracle.




Wednesday, August 17, 2016

Cool Essbase MDX Stuff – Dimension Properties Edition

I have always liked MDX as it exposes very powerful operations for creating member sets.  It has this amazing ability to take a set of members, union it with another set of members, intersect it with yet another set of members, and then exclude members from yet another set.  This is very powerful.

Recently, I was talking with a customer about some creative uses for MDX and they told me about some cool things they had done with MDX.  Kudos to George Cooper and Esam Jaber at Gap for showing me some creative MDX.  Though this is a really cool and new-to-me technique, after doing some online searches I found that the technique isn’t new to everyone.  Both Gary Crisci and Harry Gates have blogged on it already at http://garycris.blogspot.com/2014/03/mdx-queries-to-get-parentchild-and.html and http://www.cubesavvy.com/cubesavvy-utilities-updated-mdx-capabilities, respectively.

This technique, which gives developers access to outline information that was not previously available, also seems to be lightning fast.  I decided maybe I need to write on it as well and, of course, put my own spin on it by showing how we can use this in the Dodeca Spreadsheet Management System.

So, what is it?  They are called dimension properties and can utilize the PROPERTY_EXPR function.  This function provides the ability to query for, and return, members related to a given member.  The ability to return related members, on the same row, is something that is unique to the Essbase query languages.  The Grid API, or query-by-example and used to retrieve data from Essbase in the classic add-in, Smart View, and Dodeca, cannot do it out of the box.  The Essbase Report Script language, which is frequently used for exporting data, cannot do it either.  Regardless, we have customers who have asked to how have the parent member and/or the grandparent member on the same row as a member.   In the remainder of this blog post, I will explain dimension properties, and more specifically, the PROPERTY_EXPR function.

For this post, I will use the Geography dimension of the ASOSamp.Sample database.  The Geography dimension features members from different regions of the United States including details of the Region, the State, the City, and the Postal Code.  Here is a screenshot of some members in the Geography dimension:

Geography dimension members

With this dimension, let’s suppose you would like to get outline information at the city level, or level 1 in Essbase-speak.  In addition, let’s say you would like to have the state and the region as well.   Finally, while we are at it, maybe you also want to know the generation number of the member.  With dimension properties and the PROPERTY_EXPR function, you can get all of that information in one trip to the server.  Here is the query:

Select {} on COLUMNS,
Hierarchize(Descendants(Geography, Geography.Levels(1), SELF_AND_BEFORE), POST)
DIMENSION PROPERTIES
 GEN_NUMBER,
  PROPERTY_EXPR(Geography, MEMBER_NAME, Ancestor(CurrentAxisMember(), Geography.Generations(2)),"Gen2"),
  PROPERTY_EXPR(Geography, MEMBER_NAME, Ancestor(CurrentAxisMember(), Geography.Generations(3)),"Gen3")
 ON ROWS
FROM ASOSamp.Sample

Let’s break down the query to examine the components.  First, the Column axis specification is simple:

Select {} on COLUMNS,

The Column axis specification contains an empty set, but why do you need to even specify a set at all?  The Essbase MDX specification states that a query cannot skip axes based on a set order.  It seems much more natural for me to get data back on the second axis, or the Row axis, and as the Column axis is the first specified axis, a query cannot skip the Column axis and specify a Row axis.  As we really aren’t looking for any data to be returned in this query, then we can just use an empty set.

Next, let’s look at the Row axis specification:

Hierarchize(Descendants(Geography, Geography.Levels(1), SELF_AND_BEFORE), POST)

Let’s split this into sections starting from the inside out.  The Descendants function returns, naturally, all of the descendants of a given member down to, and including, level 1 members in the outline.  For this database, this query will return the Geography dimension down to the City level, but will not return the bottom, or zip code, level.

The member set returned by the Descendants function is then sorted by the Hierarchize function.

Hierarchize(Descendants(Geography, Geography.Levels(1), SELF_AND_BEFORE), POST)

The POST argument specifies that child members are sorted before their parent as they are in the spreadsheet add-ins.

Next, the DIMENSION PROPERTIES modifier for the Row axis specification provides the ability to return additional outline information related to each member returned in the set.  The DIMENSION PROPERTIES specify that the generation number, the related generation 2 and generation 3 members are returned for each member in the set.

DIMENSION PROPERTIES
 GEN_NUMBER,
  PROPERTY_EXPR(Geography, MEMBER_NAME, Ancestor(CurrentAxisMember(), Geography.Generations(2)),"Gen2"),
  PROPERTY_EXPR(Geography, MEMBER_NAME, Ancestor(CurrentAxisMember(), Geography.Generations(3)),"Gen3")

The GEN_NUMBER argument is self-explanatory, so let’s look at the PROPERTY_EXPR function to see how it works.  Consider this portion of the statement:

PROPERTY_EXPR(Geography, MEMBER_NAME, Ancestor(CurrentAxisMember(), Geography.Generations(2)),"Gen2")

The first argument is the dimension name, so that is easy enough.

The second argument, called the property_name argument, has a number of valid values as documented by Oracle in the documentation.  Valid values are MEMBER_NAME, MEMBER_ALIAS, LEVEL_NUMBER, GEN_NUMBER, IS_EXPENSE, COMMENTS, RELATIONAL_DESCENDANTS, MEMBER_UNIQUE_NAME), an attribute dimension name, an alias-table name, or a UDA.  At the time of this writing, the current docs are located at http://docs.oracle.com/cd/E26232_01/doc.11122/esb_tech_ref/frameset.htm?mdx_property_expressions.html.

The third argument is called member_value_expression and this is where the magic happens.  In this argument, you can use a number of functions that return exactly one member that is related to the current member.  In my example above, the Ancestor function is being used to return, for each member in the set as specified by the CurrentAxisMember function, the ancestor of that member at a given generation.  Among other functions are functions that allow you to get the parent, first child, next sibling, or previous sibling of the current member.

The fourth and final argument is used to give a title to the column containing the extended information.

So, now that we know a bit about this syntax, what does it look like in EAS?  

MDX dimension properties in EAS
In their blogs, both Gary and Harry talk about ways to actually consume this information.  Harry wrote a special interface to display dimension properties returned by MDX.  Gary talked about the idea of using an Excel macro to parse the output.  I had my own ideas on how to use dimension properties in Dodeca.  I setup a simple Dodeca report and used a only 3 lines of our automation language, workbook scripting, to build the view.  Here is screenshot of my simple Dodeca view.

Dynamic MDX View in Dodeca
While running this view in Dodeca, the user can filter based on Geography, Product, and Stores dimensions and the report is highly dynamic.  The automation in Dodeca performs these tasks:
  1. Returns members and properties as the descendants of a user-selected Geography member
  2. Retrieves and places the dimension properties and the members on the worksheet
  3. Places the selected Product and Stores dimension members in the proper location
  4. Retrieves Essbase data into the worksheet
  5. Creates Excel grouping based on the generation number
I will leave the step-by-step of creating this view in Dodeca to another blog post.

So, how could you use dimension properties?


Sunday, August 7, 2016

ODTUG Leadership Program / ODTUG Board Elections

As many of you know, I have spent many years helping the Hyperion community in many ways.  I have wrote this blog for quite some time, have taken on the OlapUnderground Utilities and provided free support through my company,  In other words, I have worked hard to give back to the community and, in return, have had many kind words from people we have helped.

As a result of this spirit of giving back, I was lucky enough to be elected to the ODTUG Board of Directors which brings me to the real point of this blog post.  There is an upcoming ODTUG Board Election coming up and, due to term limits, my time on the board is coming to a close.  The same is true of my good friend Cameron Lackpour.  With two board seats open, could *you* be the next person to step up?

Here are my thoughts on what it takes to be a board member.  Joining the ODTUG Board isn't just something you decide to do and it isn't an achievement that is 'a feather in someone's hat'.  It is a commitment to helping others in the community for two years.  But it is more than that.  To be a successful ODTUG Board member, the commitment you have should be a continuation of the long-term commitment you have made to helping others.  So, if you are interested in serving on the ODTUG Board at some time in the future, the thing you should do is #GetInvolved.  If you are not already involved, the easiest way to get involved is to volunteer on the ODTUG Volunteer page.

If you are ready to step up from that point, for a number of years, ODTUG has maintained a Leadership Program to help train the next generation of users.  The application process for this year's Leadership Program is open for another week.  For more information, or to submit an application for the Leadership Program, you can get more information on the ODTUG Leadership Program page.

Tuesday, May 31, 2016

Announcing the Dodeca Spreadsheet Management System, Version 7 and the Dodeca Excel Add-In for Essbase

After 18 months of hard work, Applied OLAP is proud to announce the general availability of the Dodeca Spreadsheet Management System, version 7, and the all-new Dodeca Excel Add-In for Essbase.

The Dodeca Spreadsheet Management System provides customers the ability to automate spreadsheet functionality, reducing the risk of spreadsheet errors while increasing productivity.  It combines unprecedented ease-of-use for business users using spreadsheets for planning, budgeting, forecasting, reporting and analysis tasks.  It also provides a robust, programmable development environment enabling companies to create spreadsheet applications tailored to their specific needs.

The new Dodeca Excel Add-In for Essbase was created as a drop-in replacement for the classic Essbase add-in and is the world’s only Excel add-in focused exclusively on Essbase.  The new add-in supports the most common actions used by Essbase traditionalists, supports the corresponding VBA functions, and includes a built-in Excel ribbon.  Early adopters have also been impressed by the speed of retrieving data, commenting they found the Dodeca Excel Add-In as fast as, or even faster, than the classic Excel Add-In for Essbase.  It is supported for Excel 2010, 2013, and 2016 and for Essbase 9.3.1 and higher.




Dodeca 7 includes new features and functionality for security, selectors, logging, and enhanced workbook scripting.

The enhanced security features add the ability to more easily manage users, roles, and permissions to access a Dodeca application.  The new security features include:
  • User Management – You can now track, monitor, and control user access to a Dodeca application and more easily monitor your Dodeca user base.  This feature enables customers to control individual user access to a specific application, enable users for admin access, and manage user mapping to roles while also tracking metrics such as the first, last, and count of user logins.  This feature also logs metrics on the users system to enable Dodeca administrators to more easily support their users.



    Here is an example of the metrics stored for each user record.

  • User Roles – In addition to provisioning roles via Essbase, Microsoft Active Directory, or LDAP groups using Dodeca authentication services, you can now create your own groups directly in Dodeca and map them to users.  In addition, these new roles can be configured to be additive to the roles provided by other authentication services.  
  • Application Session Timeout – You can now prevent users from keeping a Dodeca session open indefinitely by specifying an inactivity timeout or by specifying a designated shutdown time.

The new logging features provide the ability for customers to both easily track what their users are running in the system and assist our support team if and when support is needed.  The new logging features include:
  • View Usage Logging – View usage is now automatically logged in the server.  The logs include not only identifying information for the view and the user, but also include performance information, error information, and tokens used in the view generation.

  • Client-side Request and Response XML Logging – The XML traffic traveling between the client and the server may now be logged to a directory on the client machine.  This logging expands on the Server-side Request and Response XML Logging to make it easier to gather the XML traffic for a single user.  In turn, the XML captured can be used by our support team to easily replicate transactions for our developers if and when necessary.

The selector enhancements include improvements to both view selectors and member selectors.  The improvements include:
  • View Selector Relational Hierarchy Generation – You may now populate the entire View Hierarchy, or alternatively, populate one or more branches, based on the results of a relational query.

  • View Selector Hierarchy Item Access Filters – Previously, you could control the View Hierarchies available to a given user.  This new filter provides the ability to control which view hierarchy items are presented in the view selector based on the current user’s roles.

  • Relational Treeview Point-of-View Selector List– You may now configure a hierarchy of point-of-view items based on the results of a relational query.

  • Enhanced Essbase Treeview Point-of-View Selector List Expansion and Selection Filtering – You can now filter Essbase member content and selectable status based on one or more specified filters including generation number, level number, member name, alias, or shared status.
View editing enhancements make it even easier to create and modify Views in Dodeca:
  • Enhanced SQLPassthroughDataSet Query Editing – You can now define token values to use for testing tokenized SQL queries in the Test Data Set utility.
  • Improved Token Editor – You can now view and edit tokens and their values in an improved grid layout.

Workbook scripting functionality adds more flexibility to Dodeca via 108 events that provide the opportunity for customers to extend Dodeca, 116 configurable methods (actions that can be taken in response to the events), and 138 functions that provide extended information to the workbook script.

The new workbook script functionality includes:
  • New Events
    • BeforeBuildExecute - Allows a workbook script to cancel the build before the view is covered.
    • BeforeRefreshExecute - Allows a workbook script to cancel the refresh before the view is covered.
    • Shown - Raised when the view is initially shown.  The event is raised before the framework applies the view’s AutoBuildOnOpen property, which allows a workbook script to set the property dynamically.
  • New Methods
    • ExportToExcel – Provides the ability to export view and point-of-view information in an exported Excel file.  This information is used to regenerate the view upon Excel file import which enables off-line data entry in Excel with subsequent database updates in Dodeca.
    • SetSelectorConfiguration - Provides the ability to add or remove a point-of-view selector to/from a view dynamically.
  • Enhanced Methods
    • CallWebService – Added a new RESTRequest overload, which allows web service calls to made to RESTful web services.
    • SendEmail - Added a new ServletSMTP overload, which sends email from the Dodeca server rather than from the client.  This is useful in circumstances in which SMTP mail must come from an approved IP address.
    • SetEntry – Added a new Added FormulaArray overload which allows Excel array formulas to be entered programmatically.
    • SetFill - Added a new Clear overload, which clears the fill color and pattern from the specified range.
  • New Functions
    • ColumnWidth - Returns the column width based on the active cell or a given cell.
    • RowHeight - Returns the row height based on the active cell or a given cell.
    • DataPointHasCellNote - Returns a boolean indicating if the active or specified data cell has Essbase LRO cell notes associated with it.
    • IsInCharacterRange - Returns a boolean indicating whether the specified string is limited to the specified character range.  This function can be used to detect multi-byte characters in a string.
  • Enhanced Functions
    • SheetCount - Added an optional IncludeHiddenSheets argument, which controls whether the returned count includes both visible and hidden sheets or only visible sheets.
For more information, we recommend you download and read the latest Dodeca release notes from the registered section of our website.  As always, you may also reach out to us via email or call us at
256.885.4371.

Wednesday, January 6, 2016

Next Generation Outline Extractor 2.0.5.1073 released



In the last week or so, we placed an updated version of the Next Generation Outline Extractor on our website. This version provides support for some updated Essbase versions, including 11.1.2.4.002, 11.1.2.4.003, and 11.1.2.4.005. More importantly, it addresses a bug where alias names were improperly associated with parent members when using the MaxL extraction source.. This bug was reported to us by a number of users and we are glad we were able to address it. Here is a list of the issues that were addressed:

2015.11.23 - Issue 1401 - Resolved an issue where only one alias table is exported when using MaxL as the extract source.

2015.11.23 - Issue 1402 - Resolved an issue where extracts using MaxL input and having members specified with Unicode may print incorrect characters in the output.

2015.11.23 - Issue 1403 - Resolved an issue where aliases and udas may have been improperly placed on parent members.

Please contact our support team if you have any issues.

Monday, October 5, 2015

Fundamentals of SQL Writeback in Dodeca

One of the features of Dodeca is read-write functionality to SQL databases.  We often get questions as to how to write data back to a relational database, so I thought I would post a quick blog entry for our customers to reference.

This example will use a simple table structure in SQL Server though the concepts are the same when using Oracle, DB2, and most other relational databases.  The example will use a simple Dodeca connection to a JDBC database.  Here is the Dodeca SQL Connection object used for the connection.

The table I will use for this example was created with the following CREATE TABLE  statement.

CREATE TABLE [dbo].[Test](
[TestID] [int] IDENTITY(1,1) NOT NULL,
[TestCode] [nvarchar](50) NULL,
[TestName] [nvarchar](50) NULL,
  CONSTRAINT [PK_Test] PRIMARY KEY CLUSTERED 
  ([TestID] ASC)
)

First, I used the Dodeca SQL Excel View Wizard to create a simple view in Dodeca to retrieve the data into a spreadsheet.  The view, before setting up writeback capabilities, looks like this.

To make this view writeable, follow these steps.
  1. Add the appropriate SQL insert, update, or delete statements to the Dodeca SQL Passthrough Dataset object.  The values to be replaced in the SQL statement must be specified using the notation @ColumnName where ColumnName is the column name, or column alias, of the column containing the data.
  2. Add the column names of the primary key for the table to the PrimaryKey property of the SQL Passthrough DataSet object.
  3. Depending on the database used, define the column names and their respective JDBC datatypes in the Columns property of the SQL Passthrough Dataset.  This mapping is optional for SQL Server because Dodeca can obtain the required information from the Microsoft JDBC driver, however, the Oracle and DB2 JDBC drivers do not provide this information and it must be entered by the developer.
For insert, update, and delete operations, Dodeca parses the SQL statement to read the parameters that use the @ indicator and creates a JDBC prepared statement to execute the statements.  The prepared statement format is very efficient as it compiles the SQL statement once and then executes it multiple times.  Each inserted row is also passed to the server during the transaction.  The values from each row are then used in conjunction with the prepared statement to perform the operation.

Here is the completed Query definition.


Next, modify the DataSetRanges property of the Dodeca View object and, to enable insert operations, set the AllowAddRow property to True.  Note that if you added update and/or delete SQL to your SQL Passthrough Dataset object, be sure to enable those operations on the worksheet via the AllowDeleteRow and AllowModifyRow properties.

Once this step is complete, you can run the Dodeca View, add a row, and press the Save button to save the record to the relational database.



The insert, update, and delete functionalities using plain SQL statements is limited to operations on a single table.  If you need to do updates on multiple tables, you must use stored procedures to accomplish the functionality.  You can call a stored procedure in Dodeca using syntax similar to the following example:

{call sp_InsertTest(@TestCode, @TestName)}

Dodeca customers can contact support for further information at support@appliedolap.com.

Sunday, June 14, 2015

Next Generation Outline Extractor 2.0.4.887 Released

We recently released an updated version of the Next Generation Outline Extractor. This new version, 2.0.4.887, addresses three issues:


  • Fixed an issue where the username and password passed via the command line were improperly logged
  • Fixed an issue reading MaxL XML data sources when the alias or UDA contained xml encoded characters such as the ampersand (&) character.
  • Updated labels on the Input Source tab of the user interface to clarify their purpose.
Here is a screenshot showing the updated labeling.


Due to the architecture of the Oracle Essbase APIs, it is generally much faster to use the MaxL Outline XML extracts when processing an Essbase Outline extract.  The Next Generation Outline Extractor still uses the Essbase Java API during this extract, but it is able to minimize the number of calls.  The second option shown above, Extract and Process MaxL Outline XML, will automatically extract the Outline XML from the cube during the processing.  The third option shown, Use Previously Extracted MaxL Outline XML, uses (obviously) an Outline XML file that has already been extracted.

Thank you to everyone who reported issues or made suggestions as you help make this utility better!