Monthly Archives: March 2017

Using Oracle HTTP Server 12c with Oracle EPM (Hyperion) 11.1.2.4

March 29, 2017

Author: Andy Tauro, Performance Architects

Ever since Oracle released Oracle Business Intelligence Enterprise Edition (OBIEE) 12c (12.2.x), it has been a waiting game to see when Oracle EPM (Hyperion) would move to the 12c versioning.  As of the date of publication of this blog post, the latest version of Oracle EPM is 11.1.2.4.x, which bundles many improvements over the prior releases of Oracle EPM. However,  Oracle Fusion Middleware (FMW) is still 11g, to be precise 11.1.1.7. This includes Oracle HTTP Server (OHS) 11.1.1.7, and we have been asked more than once when Oracle EPM will be updated to use the latest version of FMW or OHS.

As a workaround to meet Infrastructure standards, we often set up a standalone instance of OHS with EPM. Until recently, the version of OHS that we typically used was OHS 11.1.1.9, which is the latest version of OHS in the 11g family.

With the migration to 12c versions of Oracle Relational Database (RDBMS), OBIEE and WebLogic, the 12c version of OHS is also being adopted as part of some of these products. The natural progression is to bring EPM into the fold. But the transition is not simply plug-and-play.

OHS 12c is based on Apache HTTP Server 2.4, whereas the 11g version was based on 2.2. The change in the base code results in changes in the implementation. While most of the functionality is the same, there are some key differences in the directives used. For instance, “Order” and “Deny” directives have been replaced by the “Require” directive. Luckily, there is an Apache webpage that lists a lot of the changes:  https://httpd.apache.org/docs/2.4/upgrading.html

This is not a certification of Hyperion EPM 11.1.2.4 with OHS 12c. While we have seen it work for the uses that we have put this combination through, this does not imply Oracle’s support for the combination at this time. The reason for this is most likely the fact that Oracle EPM is closely integrated into the instance of OHS that gets deployed with it. In the above-described scenario, the 12c instance of OHS serves as a duplicate to the bundled instance and serves as a proxy to the requests that it serves.

While there are benefits to integrating OHS 12c with Oracle EPM 11.1.2.4, such as updated security code and better performance, it is necessary to keep in mind the Oracle Support certification of the solution. Continued support from Oracle for a complex suite of tools like Oracle EPM is essential, so keeping the solution simple helps. And this also reduces the chances of the solution breaking down when the next upgrade comes around.


© Performance Architects, Inc. and Performance Architects Blog, 2006 - present. Unauthorized use and/or duplication of this material without express and written permission from this blog's author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Performance Architects, Inc. and Performance Architects Blog with appropriate and specific direction to the original content.

The Performance Architects Insider’s Guide to COLLABORATE 17

March 27, 2017

Author: Kirby Lunger, Performance Architects

COLLABORATE 17, the largest meeting of Oracle user groups representing the full family of Oracle business applications and database software, is taking place Sunday, April 2nd through Thursday, April 6th at Mandalay Bay in Las Vegas. The event is co-hosted by the Independent Oracle Users Group (IOUG), the Oracle Applications Users Group (OAUG), and Quest International Users Group (Quest), which means this is the biggest meeting of Oracle users outside of Oracle OpenWorld!  As a result, the event can be somewhat difficult to navigate, so the Performance Architects team wanted to share our thoughts on important happenings at the conference for the BI and EPM (Hyperion) communities.

Pre-Conference Workshops

The OAUG holds pre-conference workshops on Sunday that provide “deep dives” into specific topics; these are a great way to get detailed instruction (and sometimes continuing professional education credits!) and to network with like-minded people.  Most of these cost an additional fee to attend.  The Performance Architects is hosting one workshop:

Special Interest Group Sessions and Events

There are four key special interest group (SIG) sessions and events that you should add to your calendar for BI/EPM-related topics:

  • Monday, April 3rd, 11:00 AM–12:00 PM: OAUG Hyperion SIG Session (Session ID: 10178 – Breakers E). Register here: https://app.attendcollaborate.com/event/member/317668
  • Tuesday, April 4th, 4:15 PM–5:15 PM: OAUG Endeca SIG Session (Session ID: 10010 – Breakers L). Register here: https://app.attendcollaborate.com/event/member/317501
  • Wednesday, April 5th, 11:00 AM – 12:00 PM: OAUG Oracle Business Intelligence – Big Data SIG Session (Session ID: 10214 – Breakers E). Register here: https://app.attendcollaborate.com/event/member/317704
  • Hyperion Connect Reception! Join us on Tuesday, April 4, from 7 p.m. – 8:30 p.m. and network with your fellow Oracle EPM users at a special reception just for this community within COLLABORATE. Share ideas, make connections and meet Oracle executives and product developers. The reception will start at 7 p.m. in the Palm Foyer of the North Convention Center and preregistration is not required!

Presentations

There are 1200+ sessions organized into 17 tracks, so it’s essential to figure out a schedule during the conference (just getting to sessions during the breaks can be tricky!).

We strongly recommend attending the Oracle roadmap sessions, as these are probably the only public roadmap sessions you’ll get to see until Kscope this June or OpenWorld in the fall.  For EPM, we recommend:

For BI, there aren’t any “official” roadmap sessions, so we recommend the BI-Big Data and Endeca SIG meetings mentioned above, as we know Oracle product management leaders will be attending both SIGs and presenting on product strategy and roadmap details at those sessions.

We would also be remiss if we didn’t mention that the Performance Architects team has several presentations scheduled during the conference:

Fun and Networking

The conference hosts some really fun events so that you can meet your peers and enjoy their company!  The Welcome Reception is held on Sunday night at the House of Blues and is a “don’t miss” way to see who’s attending the show this year.  The Monday Welcome Reception in the Vendor Showcase is also a fun kickoff event (with free drinks and food) and is a helpful way to see who’s exhibiting and what’s new in the vendor world (Performance Architects will be hosting a fun game at Booth 847).  If you’re lucky enough to get to stay for the end of the conference, the Attendee Appreciation Party on Wednesday night should also be on your schedule.

Performance Architects is also hosting a special, invite-only event during the conference.  If you’re interested in learning more, please email us here and we’ll let you know if there’s still space to participate.

We look forward to seeing you soon!

 


© Performance Architects, Inc. and Performance Architects Blog, 2006 - present. Unauthorized use and/or duplication of this material without express and written permission from this blog's author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Performance Architects, Inc. and Performance Architects Blog with appropriate and specific direction to the original content.

Clearing Data in the Oracle Planning and Budgeting Cloud Service (PBCS) Simplified Interface

March 22, 2017

Author: Ben Hogle, Performance Architects

If you’ve ever loaded data into Oracle Planning and Budgeting Cloud Service (PBCS), you’ve almost inevitably had to clear data from your ASO and/or BSO cubes. Maybe you goofed, maybe something changed; it doesn’t matter, your data must be purged and reloaded; and the ability to clear large amounts of data relatively easy is a life saver for many of us in this business. This is generally a simple process that can be completed with a few simple steps.

Within the system, users can clear numerical cell data as well as additional cell details such as annotations and supporting details.  We’ll focus only on clearing numerical data in this blog post.

As most of you likely know by now, Oracle will discontinue use of the Standard Interface late this calendar year so I will only show how to do these tasks in the Simplified Interface (Need more information on this topic?  Check out the Performance Architects blog post: Oracle Planning and Budgeting Cloud Service (PBCS): Twelve Month Farewell Tour for the “Standard Interface”).

Clearing Data from cubes in the Simplified Interface can be very simple and easy by following a few quick steps:

  • From the home screen click: “Application>Overview”

  • Near the top right-hand side click: “Actions>Clear Cube”

  • Click: “Create”

  • Name your job and select the cube to be cleared from the drop down list, then select what data to clear.
    • The choices for a BSO cube clear are:
      • Clear All – Delete all data
      • Upper-Level Blocks – Delete data in upper level blocks only
      • Non-Input Blocks – Delete data in non-input blocks only
      • Dynamic Blocks – Delete dynamic block data only
      • Empty Blocks – Delete only blocks with values of: #MISSING
  • An ASO cube clear offers fewer options:
    • Clear All – Delete all data
    • All Aggregations – Delete all aggregated values
    • Partial Data – Clear data from a specific location (i.e., data attached to a specified scenario, version, etc.); Partial Data clear also lets the user decide whether they want a logical clear (essentially zeroing out the numbers) or a physical clear (actually removing all values from each cell)
  • Directly below are the BSO cube clear options and second below are for the ASO cube clear.

BSO

ASO

  • When finished select “Save and Close.” After the job is saved, you can then run the action

When you have finished adding and running the action, you’ll want to go through the data and verify that it has all been cleared from the proper cube that you selected. To do this, simply go to any created forms or reports that have previously been generated and verify that no data exists in the cube(s) that were cleared.

Several other ways to clear data exist, but I don’t have enough time in one blog entry to discuss them all. Please feel free to reach out us at communications@performancearchitects.com if you have any questions about clearing data in the PBCS Simplified Interface.


© Performance Architects, Inc. and Performance Architects Blog, 2006 - present. Unauthorized use and/or duplication of this material without express and written permission from this blog's author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Performance Architects, Inc. and Performance Architects Blog with appropriate and specific direction to the original content.

Important Patch Set Exceptions (PSEs) for Hyperion Planning 11.1.2.4.005

March 15, 2017

 

Author: Andy Tauro, Performance Architects

For folks who are deploying Hyperion Planning (the Oracle EPM on-premise budgeting, forecasting and planning solution) 11.1.2.4.005, or for those of you who are already on that version, the Performance Architects team strongly recommends to apply Patch Set Exception (PSE) 24597360.

One very important consideration: PSEs are usually not compatible with each other. If you already have a PSE applied for 11.1.2.4.005, you may run into conflict and have to choose which one to keep. Patch Set Updates (PSU) do not usually have this problem because they are cumulative starting 11.1.2.4 (except for Shared Services) and include all fixes issued before them.

This PSE fixes two pretty important bugs with data entry web forms.

  1. Account dimension fixes

In most uses of web forms, the Account dimension is in the row. However, in cases where one or more of these members are also present in an alternate hierarchy, the row that the member is in does not show in the row labels, but the actual data row for the member does show up, especially when the setting to “Suppress missing blocks” is enabled. As a result, the data form gets skewed at best, and can get completely unusable unless this PSE is applied.

  1. Columns widths

Without the PSE, column widths do not respect width settings and instead resize to the width of the column labels. This can be very annoying for users, even if it does not make the web form unusable. For instance, if the header of a column was ‘FY17’, the column width would essentially be four characters wide. Not just would that make for a very slim column that would need to be manually resized to read a number larger than four digits, because all the column headers are most likely not the same, the form will look very ugly.

Please note that this PSE does not completely fix this problem.  On first opening the web forms, the column widths will be sized as per design settings. If any columns are set to be able to expand (such as ‘YearTotal’), the additional columns are squished as well to the size of, for example, ‘Qtr1.’ However, refreshing the web form at that point fixes the additional columns. There is a bug pending with Oracle Development for this (24838130).

If you are seeing any of these issues in your environment, hopefully this note helps you. If you feel that you will require expert assistance in applying this PSE to your environment, or with something else in on-premise Hyperion Planning or Oracle Planning and Budgeting Cloud Service (PBCS), please feel free to drop a note and we will be in touch.

 

 

 


© Performance Architects, Inc. and Performance Architects Blog, 2006 - present. Unauthorized use and/or duplication of this material without express and written permission from this blog's author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Performance Architects, Inc. and Performance Architects Blog with appropriate and specific direction to the original content.

How to Implement Slowly Changing Dimensions (SCD) Type 2 in Oracle Data Integrator (ODI) 12c

March 2, 2017

Author: Margaret Motsi, Performance Architects

Tracking historical changes in data (slowly changing dimensions) is a very common Oracle Data Integrator (ODI) task since many industries require the ability to monitor changes and to be able to report on historical data accurately at a point in time. Implementing slowly changing dimensions (SCD) in ODI 12c is relatively easier than in 11g.

This blog post provides instructions on how to implement SCD in 12c, and points out any of the differences from 11g.

SCD Target Table Requirements

In the data warehouse (same as in 11g), the target table needs to contain these additional columns:

  • A natural key. This represents the primary or unique key, usually from the source system.
  • A surrogate key. This should be a number unique for each data entry that is automatically generated, usually using a sequence for example the ROW_WID column it only exists in the warehouse and is based on a DB or ODI sequence.
  • A starting timestamp. Indicates what date the record is created in the data warehouse.
  • An ending timestamp. Indicates what date the record changes.
  • A current record flag. Indicates whether the record is the current one (1) or the historical one (0).

In this example, the following are the additional columns:

  • Natural Key: EMPLOYEE_ID
  • Surrogate Key: ROW_WID
  • Starting Timestamp: EFF_FROM_DT
  • Starting Timestamp: EFF_TO_DT
  • Current Record Flag: CURRENT_FLG

Configuring the SCD Mapping

  1. Create the target data store model in the Designer and define the SCD behavior for each column.

In ODI 11g this was a tedious exercise, as the column editor had to be opened for each column one-by-one, then defined in the SCD Behavior area in the Description tab. In 12c, all of the columns can be defined at the same time in the Model Editor under the Attributes tab.

ODI 11g

ODI 12c 

  1. In the Projects section, create the mapping to load the SCD target table.

ODI 12c

Another thing to note is that in ODI 12c, the mapping editor has been simplified to only three tabs instead of the seven in 11g (see below). The Mapping tab is now called Logical and Quick Edit can be accessed within this area. The Flow tab is now called Physical, and Flow Control can be accessed in this area.

ODI 11g

  1. After dragging and auto-mapping the source and target tables, map the SCD specific columns as follows:
  • EFF_FROM_DT. Maps to ‘SYSDATE’ which captures the date when record is created.
  • EFF_TO_DT. Maps to ‘SYSDATE’ which captures the date the record is changed. The default value from the SCD IKM is 01-01-2400.
  • Maps to ‘Y’. The KM populates this column so no need to populate this column. (1: Current Record, 0: Past Records).
  • ROW_WID. Maps to a sequence which auto-populates unique values for each record. In this case, we are mapping to a ODI sequence.

As mentioned earlier, in ODI 12c the Quick Edit tab no longer exists. Now, you can easily access the Quick Edit panel from the mapping by clicking on the target data store and by viewing the Attributes panel in the Property Inspector as shown above.

The Execute On parameter in 11g has been replaced with the Execute on Hint parameter in 12c, which can easily be specified in this same panel, as shown below.

ODI 11g

ODI 12c

  1. The only step that needs to be done in 12c which was not necessary in 11g is to define Integration Type in the target table property inspector under the Target tab.

Lastly the Integration Knowledge Module (IKM) is set in the Physical tab (formerly the Flow tab in 11g). As mentioned above, the Control tab is no longer there in 12c and Check Knowledge Modules can be defined under the Physical tab in the Property Inspector.

Quick tip: If the Primary Key of the target table is indexed, make sure it is non-unique as the historized data will create duplicate primary keys.

Once the mapping is tested, similar to 11g, it can be added to a package and then will generate a scenario for that package. This will enable data loads to capture any changes and will historicize the old data whenever the scenario is executed. In the below example, the JOB_ID for four employees changed when they were promoted to managers and this is captured in the data load.

Still have questions about implementing SCD in ODI 12c?  Contact the Performance Architects team here and we would be happy to help.


© Performance Architects, Inc. and Performance Architects Blog, 2006 - present. Unauthorized use and/or duplication of this material without express and written permission from this blog's author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Performance Architects, Inc. and Performance Architects Blog with appropriate and specific direction to the original content.