Tim Tow

As promised, I am starting a series on the Collaborate 08 presentation by Aneel Shenker that discussed new features scheduled for the next version of Essbase. This presentation, as any presentation you see from Oracle on unshipped product, has the following disclaimer:

"The following is intended to outline our general product direction. It is intended for information
purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle."

The fact that you hear it from me certainly doesn't change the above disclaimer as I can't/don't speak for Oracle.

That being said, here is a list of the things presented with a brief description of benefits of these features:

1. Varying Attributes (aka Slowly Changing Attributes) - Slowly changing attributes let you change how members rollup based on time. A common usage scenario would be in a corporate reorg where 'Dept 100' rolls up to 'Mgr A' until July 2007, then it rolls up to 'Mgr B'. Without slowly changing attributes, a 'reorg' in the Essbase outline moves all history to roll up to 'Mgr B'.
2. Time Intelligence enhancements in block storage. These were introduced in 9.3 but were ASO only; now some of that same functionality will be available throughout Essbase.
3. Date and Text Measures - Allows you to display text instead of a value as the result of an Essbase query. To me, it looks like an enhanced 'Missing Text String' type of functionality except the 'text' values are controlled by the database designer.
4. Improvements to backup and logging. The thing that stands out for me on this is the ability to audit logging capabilities. It appears Essbase will now track all write operations to the database including outline changes, data load, calculations and lock and send. Further, there will be new 'replay' feature that lets you 'replay' these 'writes' to your database; it looks very useful in the case where you have to restore from a backup.
5. ASO Improvements including the ability to do a 'Partial Clear' of the database and ability for ASO databases to be the target of a partition.
6. Essbase can be a physical database source in the OBIEE BI Server. For those of you that have always wanted to use SQL to go against Essbase, essentially the BI Server layer will allow you to do that; I would guess they are putting a virtual star-schema layer on top of Essbase in the BI Server layer.
7. XOLAP or 'eXtending OLAP on a relational database'. Essentially, XOLAP is dynamically sourcing the cube from a relational datasource. I think this could be a very exciting development that could allow us to write some very cool applications in the future that were difficult before. For example, payroll planning applications where the user wanted to add, on the fly, new employees to plan for may be much easier in XOLAP. I can't wait to get my hands on it so I can see how far we can push it.

The presentation also talked about Essbase Studio and some future Essbase things as well. I plan to post a bit more on this topic and include some screenshots as time allows.

If you want to learn more about Essbase 11.1.1, I would urge you to attend the ODTUG Kaleidoscope conference in New Orleans. The Essbase Symposium on Sunday, June 15 is dedicated exclusively to 11.1.1 and future versions of Essbase; speakers for the day are all from the Essbase development organization and it will give you the unique opportunity to get your questions answered directly from the developers. Information on Kaleidoscope is located at http://www.odtugkaleidoscope.com/hyperion.html.

Sorry I have been so busy and not able to post a play-by-play account of Collaborate. My friend Edward Roske did manage to post often from Collaborate and I recommend you read his comments at http://looksmarter.blogspot.com/. I don't know how he kept up; he must have been blogging the entire time he wasn't presenting..

As for my overall impression.. It does seem to me, and to others I talked with, like this is the replacement for the Hyperion Solutions conferences of the past. It did seem a bit light on 'how customer A implemented Hyperion product XYZ' types of presentations but hopefully there will be more submissions in future years. Edward Roske and Tracy McMullen from interRel did many, many presentations; they did invest a bunch of time in submitting/writing many of the presentations. The best presentation I saw was, however, a non-technical presentation by John Scott of interRel about the good, bad and ugly of budget and forecasting systems. John used to be the Global Domain Lead for Hyperion Planning and talked not only about how budgets can be good, but also on how they can lead to 'bad' behavior such as causing people to protect their budget dollars even when the business could be optimized by a lower budget..

The OAUG SIG meeting was interesting; Edward wrote about many of the details but there is one thing I wanted to point out about the benefits of SIG membership. Oracle E-Business Suites customers were facing a deadline on upgrading a specific version before it went to 'Sustaining Support' which is the lowest level on their support matrix; the SIG convinced Oracle to extend that support by (at-least) one year. The lesson? Those of you on Essbase 7X may want to join the OAUG SIG 'in-force' and together work to make sure 7X does not go to 'Sustaining Support'.

On the Essbase roadmap topic, there was an excellent presentation made by Aneel Shenker; he was gracious enough to give me a copy of the slides and I will post a my thoughts of the new features coming in Essbase 11.1.1, formerly known as "Kennedy" and "Essbase 9.5" next week. That being said, if you want to know the real scoop on this release, the ODTUG Kaleidoscope conference in New Orleans will dedicate the entire day Sunday, June 15, to the topic and will feature the Hyperion developers who wrote the product and know it through and through. I encourage you all to register for this conference soon as there is limited space available. Information on this conference is at http://www.odtugkaleidoscope.com/hyperion.html.

There was a comment on one of my earlier posts asking the status of the Advanced Security Manager. My developer is still working on the code although it is not on the top of his priority list. After a brief discussion, it appears the code changes, as requested by the OLAP Underground team as a condition for releasing the code, is complete and he has been working on properly commenting the code. I have directed him to make sure our automated build processes for ASM are in place and, once that is done, I would like to have some users start testing. If you are interested in testing the ASM, please drop me an email at timtow@appliedolap.com; it will probably be a week or two before I get anything out.

I also intended to blog 'real-time' about the Collaborate conference. Unfortunately, I didn't even have time to eat most of the time I was there. I will try to post something soon (after a few Dodeca installs this week, reviewing a couple of patent filings, doing a bunch of followup calls, etc).

I am leaving this afternoon to attend Collaborate. Actually, that isn't quite true as I am skiing in Vail tomorrow before heading back to Denver for Collaborate. In any case, I was asked to help spread the work about the Hyperion Special Interest Group meeting on Sunday afternoon. I will be there as I have been asked to serve as a 'Domain Lead' for Essbase (along with my friend Edward Roske). Here is the agenda:

• 1:00 – 1:15 OAUG Hyperion SIG Welcome and Introductions


• SIG Board members
  
• Domain Leads
  
• Web Cast Series


• 1:15 – 2:15 Oracle Key Note Speaker: Fred Richards, Oracle – Senior Director, BI Product Marketing


• 2:15 – 2:30 OAUG Board Member Speaker: Ray Payne, OAUG - Vice President, Global Outreach & Transition


• 2:30 – 2:45 Collaborate Hyperion Presentations Overview


• 2:45 – 3:15 SIG website overview and Q& A for ideas


• 3:15 – 4:15 Cocktail & Appetizer Reception

Anyone attending Collaborate who is interested in Hyperion technology is invited to attend. I hope to see you there.

I have been working for about a week on making our Dodeca product work on DB2. I finally got it done today. First let me explain a little about how Dodeca works to set the context for my post.

Dodeca runs on metadata that describes every part of the application. All report definitions, Essbase connections, point of view selectors, toolbars and even custom code are stored in a relational database and are retrieved to the client via web services on 'as-needed' basis. The system is so completely run by metadata that an application can literally be exported to a single zip file. We often create prototypes for customers, ship them an exported zip of the metadata and, within minutes, they have their prototype up and running on their system.

On the middle tier, we use an open source Object-Relational Mapping ("ORM") technology called Hibernate. Hibernate takes care of persisting the Plain Old Java Object ("POJO") representations of the metadata in our server to a relational database. Hibernate uses a 'Dialect' to create the SQL syntax necessary for a particular database flavor. We have used Hibernate with a number of databases but most frequently with MySQL, SQL Server and Oracle. We have been able to get each of these databases running with a minimum of effort. My guess is that not as many people have used Hibernate with DB2 or that they haven't used it like we are using it.

In particular, we are storing the key metadata field in a Character Large OBject ("CLOB") field. Support for CLOB data among the relational database vendors is quite varied; DB2 does something quite different. After some research, I found that the CLOB fields produced by our Hibernate schema generation were very small. In fact, they were defaulting to a size of 255 bytes. That is pretty small for any piece of metadata, particularly an Excel file. I also found that DB2 CLOBs used a number and a scaling factor in their DDL language like this:

ENCODED_ARTIFACT CLOB (1024 M )

I looked at the Dialect base class in Hibernate and found an interesting method to override:

public String getTypeName(int code, int length, int precision, int scale) throws HibernateException

Based on my experience with Hibernate over the past couple of years, I guessed that the arguments passed to this function were obtained from the Hibernate mapping file. I tested that assumption with the following mapping:

<property name="EncodedArtifact" column="ENCODED_ARTIFACT" type="text" length="16" scale="2">

I was correct and decided to use the scale option to convert to int passed in to the method to a string in the format that DB2 could understand. In fact, I decided to let Hibernate create the initial string for a CLOB type and then I would modify it to fit my needs. I wrote the following class and used my new Dialect, com.appliedolap.dodeca.hibernate.dialect.DB2Dialect, in my hibernate.properties file.

Here is the Java code for the dialect:

package com.appliedolap.dodeca.hibernate.dialect; 
 
import org.hibernate.HibernateException; 
import java.sql.Types; 
 
/** 
 * An SQL dialect for DB2 with fixes for Clob column creation for Dodeca. 
 * @author Tim Tow 
 */ 
public class DB2Dialect extends org.hibernate.dialect.DB2Dialect { 
 
    public DB2Dialect() { 
        super(); 
    } 
 
    /** 
     * Get the name of the database type associated with the given 
     * {@link java.sql.Types} typecode with the given storage specification 
     * parameters.  Note: this code calls the parent class getTypeName method 
     * unless the type is a CLOB.  If the type is a CLOB, the length field from 
     * the Hibernate mapping is used as the length of the CLOB.  Valid values are 
     * from 1 to 99999. 
     * 
     * The scale field is used to specify the scale of the CLOB.  Valid values are 0 - bytes, 1 - kilobytes 
     * 2 - megabytes, and 3 - gigabytes.  The default is megabytes. 
     * 
     * @param code The {@link java.sql.Types} typecode 
     * @param length The datatype length 
     * @param precision The datatype precision 
     * @param scale The datatype scale 
     * @return the database type name 
     * @throws org.hibernate.HibernateException If no mapping was specified for that type. 
     */ 
    public String getTypeName(int code, int length, int precision, int scale) throws HibernateException { 
        // validate the length 
        if (length < 1 || length > 99999) { 
            length = 1024; 
        } 
 
        // seed the typename 
        String result = super.getTypeName( code, length, precision, scale ); 
 
        // fix up the clob type for DB2 to create it properly 
        if (code == Types.CLOB) { 
            StringBuffer buffer = new StringBuffer(16); 
 
            // add the first part of the string 
            buffer.append(result.substring(0, result.indexOf(")"))); 
 
            // add the scale for bytes/kbytes/mbytes/gbytes 
            switch (scale) { 
                case 0: 
                    buffer.append(" "); 
 
                    break; 
                case 1: 
                    buffer.append(" K "); 
 
                    break; 
                case 2: 
                    buffer.append(" M "); 
 
                    break; 
                case 3: 
                    buffer.append(" G "); 
 
                    break; 
                default: 
                    buffer.append(" M "); 
 
                    break; 
            } 
 
            buffer.append(")"); 
 
            result = buffer.toString(); 
        } 
 
        return result; 
    } 
} 

After a bunch of work, the ODTUG agenda is finally posted and if you want to keep up on the latest coming in the Essbase world and have access to the people who know, this is the place to be. Here is a link to the conference agenda:

http://www.odtugkaleidoscope.com/hyperion.html

My friend Edward Roske has already blogged about the trial and tribulations of putting together the track. His post, "Best Essbase Conference Ever?", is a good read. Here is the link:

http://looksmarter.blogspot.com/2008/03/best-essbase-conference-ever.html

Finally, there is a Hyperion-specific registration page that has just been published by the ODTUG team at:

http://www.prairiesystemsgroup.com/odtug_hyperion_registration.html

Hopefully I will get some technical content posted here again in the next few weeks. I am off to Denver the week before Collaborate (again in Denver) for OBIEE training; I hope to post on some of what I learn there.

Until then, back to writing a custom Hibernate Dialect for DB2 (the as-shipped Hibernate DB2 Dialect doesn't create CLOB fields properly nor does it properly write a where clause if the column is an integer.. Arghhh!)

I have an update on the ODTUG Conference that I wanted to share with everyone. Edward Roske and I have been working hard on putting together a killer program that will attract people who are interested in learning, in-depth, what there is to know about Essbase. Edward has proposed a session schedule that just blows me away. It will be very unlike the Solutions Conferences over the past few years that really focused on customer success stories along with a moderated dose of 'what has recently shipped'. Rather, this conference will be all technical, period. We hope to get the sessions nailed down this week so we can make the agenda public.

Another thing I like about this conference is that, instead of the traditional golf outings that *used* to be a fixture at Solutions (years ago), this conference has a community service event on Saturday. Although the French Quarter and the Garden Districts of New Orleans have recovered relatively well, much of New Orleans is still reeling from the effects of Katrina. I haven't had the opportunity to get back to New Orleans since flying a couple of AngelFlight missions just after the storm and I look forward to helping again.

If you are an Essbase developer, mark your calendars for June 15-19 for New Orleans. Keep an eye on the website at http://www.odtugkaleidoscope.com/ for upcoming Hyperion announcements.

Sorry I haven't had time to post anything for the last couple of weeks. I plan to start a series on how to install the Hyperion 9.3.1 Suite when I can get enough time to do the proper job of it. I have been working 'half-days' (12 hours, get it?), however, for the past couple of months and haven't had much time for anything. I did, however, force myself to go fly my airplane on Saturday. You know it hasn't been flying enough when, for the second time in a row, the battery was low and the plane needed to get a jump start. It had been a couple of weeks since it had been flown and I thought I had better make sure the battery was charged up before I fly over to Atlanta on Wednesday (tentatively) to go to the Hyperion User Group meeting. I bet you didn't know you could jump-start an airplane, did you?

I did, however, get an interersting email over the weekend from some of the officers in the the Oracle Developer Tools User Group ("ODTUG"). They have decided to pursue a Hyperion technology track at the annual Kaleidoscope 2008 conference scheduled for New Orleans June 15 to June 19. Hyperion-oriented people may want to attend the conference to see the latest on both Essbase technologies and OBIEE. This conference appears as though it will have content that has been been missing, for the most part, from the Hyperion 'Solutions' conferences over the past few years where there seemed to be slanted a bit more towards marketing than technical. Information for the Kaleidoscope conference can be found at

http://www.odtugkaleidoscope.com/


Although submissions for the ODTUG conference have been closed for a while, they just this past weekend reopened submissions for Hyperion related content. Submissions are scheduled to end February 16th. You can enter your submissions at

http://www.prairiesystemsgroup.com/odtug_hyperion_abstracts.html


I hope that many Hyperion experts are willing to share their knowledge in New Orleans.