Oracle DRM Blog

Having been writing for a little while and that’s because handling a new born takes up alot of time but hopefully in the near future this blog can be expanded to cover more diverse topics. We’ll see. Got an interesting email from S.Varshne below and we’re going to try to answer it. I do take our reader’s comments very seriously and a large percent of them gets replied. Oracle DRM continues to be one of the best Data Governance tool for Financial MDM out there in my opinion and no doubt tracking changes is important to the piece of the puzzle. There’re two different ways of tracking changes.

• 1. the native DRM log – this provides very detail for every action for every node
• 2. the less detailed DRM properties tracking the date/time of the most recent changes.

Let’s review the questions…

I tried what you suggested here in baby steps
- First enabled TrackChanges System preference
- Created a property called NodeChangedFlag (You mention in your article that I can pick and choose the preferences I want to use. So I leave out the rest)
- Updated the System preference NodeChangedProperty to point to NodeChangedFlag

Here are the error messages I see…
Read the rest of this entry »

Gotten this rotten Oracle 10g error after installing the Hyperion EPM 11.1.1.2 software (without configuring it).  Anyhow, found an obscure fix for it too, apparently this is an issue with 10.1.0.2. If you happen to get this error, it is because the OS user that you are trying to login to the Oracle Enterprise Manager with has not been setup to allow the user to logon as a “Batch Job.”

To resolve this issue:

1. Go to “Control Panel” -> “Admin Tools” -> “Local Security Policy.”
2. Within “Local Policies”, go to user “Right Assignment.”
3. Add the user to “Logon as a Batch Job.”

The logon problem should now be resolved. This solution will only work for Oracle 10g on Windows 2003 Server. Have fun!

Got a question today about the new features on Oracle Hyperion DRM v11. In terms of this new release to be honest, most of the features had been pretty much the same as version 9.3.2. So don’t expect anything dramatic. I was told by reps that the next release will evolve to something quite different. Time will tell.

On with the list of new features that I find useful:

Synchonize by Property – The Synchronize feature has been extended to enable easy searching across multiple hierarchies simultaneously based on any common property value. Relationships between nodes using a property value can be discovered and navigated without creation of a query resulting in higher productivity when browsing mapped nodes across multiple hierarchies. This feature is more a visual tool.

Migration Reports – Reports can now be generated from the Migration Utility to provide offline HTML documentation for metadata extract files that are used for migration or archival. These reports can provide users the ease of reference to information stored in the XML file and allow additional user notation to be included in the report to capture the intended purpose and usage of the file. In my previous blog entry, I have opted to use the XML export in conjunction with Excel to document DRM members and properties. No doubt this new feature could come in handy.

Read the rest of this entry »

This is a little tip on Perl for those interested. Perl does not distinguish variables between text or numbers. And so when you’re manipulating data or input, you need to be careful and make your own checks first. It may seem simple but also can be way more complicated. For example, what if you have a string that combines numbers at the beginning and string at the end? What if the number is more like an IP address than a real number? What if you have negatives? Here’s a few lines of code you can use to determine the various checks you can do using regular expressions:

if (/\D/) { print “has nondigits\n” }
if (/^\d+$/) { print “is a whole number\n” }
if (/^-?\d+$/) { print “is an integer\n” }
if (/^[+-]?\d+$/) { print “is a +/- integer\n” }
if (/^-?\d+\.?\d*$/) { print “is a real number\n” }
if (/^-?(?:\d+(?:\.\d*)?|\.\d+)$/) { print “is a decimal number\n” }
if (/^([+-]?)(?=\d|\.\d)\d*(\.\d*)?([Ee]([+-]?\d+))?$/) { print “a C float\n” }

Enjoy the week guys!

Sorry folks I have not blogged for a while. And that’s because my lovely wife has given birth to our daughter and we had been quite busy and having lots of fun with her. Hopefully my schedule will also me to get back into some sort of cadence to update this. Also later on, I’ll look to expand the blog to a broader focus rather than MDM only.

While talking to many folks I came to realization that building hierarchies in DRM is still a mystery to some folks. And I apologize because I had previously attempted to address this with one of my invention which complements DRM to make hierarchy development even easier. But I probably should have gone back to the basics knowing not all our readers know how to create hierarchies using the native functionalities.

With today’s blog I just want to revisit the basic of creating a hierarchy using DRM Automator. I want to do that because I am assuming (do tell me if I am wrong) that you know how to add each parent or child member using the DRM GUI. But doing so will be extremely slow by hand. As a result, you’d want to be using the DRM Automator instead.

Read the rest of this entry »

Data Relationship Management currently allows objects (e.g. DRM property name, Exports, Property Queries, etc.) to be created using special characters. Please be aware some of these characters can cause issues when using the system. Here’s some examples:

• (comma) – Causes issues with comma-delimited lists. Besides objects, if you’re going to use Ancestor List, you will run into problems with the commas being the delimiter that will cause problems.

• <>”‘=^ – Can cause issues with the Migration utility

• (), – Can cause issues with the Formula parser

You should avoid creating object names that contain these characters where possible. As usual, just use something that’s simple such as alphanumeric plus hyphens perhaps,  and you’ll be fine.

I love Oracle to death but on some things, it bugs me. I have been trying to get somewhere with Oracle for about a month, and the issue surrounds the installation of the Excel 2007 Essbase client (not the SmartView client though). Here’s the story…

For the past 12 years we have use Essbase on Excel, users are required to download the Essbase Client for Excel. The installation up to System 9 version 9.3 is normally around 50Mb or less. That’s a reasonable size for users to download and install.

However, with the new Oracle EPM 11.1.1 and 11.1.1.1 you’ll be surprised to know that you’ll now need to tell an average “Joe the Plumber” user that he/she needs to download 1.9Gb of zip files (5 of them) and then unzip it to his/her disk drive of 3.5Gb  !!!! And then…
Read the rest of this entry »

Oracle has come out with a simple Oracle EPM 11.1.1 installation tutorial. Try it out if you’re interested in building a simple single machine install of most of your Hyperion app. I have done both an SQL 2005 and Oracle 10g install for Oracle EPM 11.1.1 on a single machine and it’s not too hard at all. It’s much simplified with the new installer which sequences the install for you…almost brainless. Check out how to install this version.

I also found another similar tutorial from Tim Tow here in PDF format. Both are quite valuable.

For production, however, you’ll definitely need to do a much more intense manual deployment.But if you’ve done a 9.3 install, you’ll know what it’s likeand what you’re up against. Some of things you need to watch out for:

1. For multiple server deployment, do you have your DNS or HOST file configured.

2. Remember to download all the right packages. If you’re installing Workspace, ensure you have the migration utility and portlet download as well.

3. If you’re hosting an Oracle DB on DHCP, you’ll also need to install the Microsoft Lookback Adaptor.

4. Make sure your DB has the right privilege. e.g. SQL 2005 requires bulkinsert privilege.

Oracle DRM 11.1 is a separate install though, you’ll not need to go through the installer. If you have done previous version 9 releases, it’s pretty much the same deal.

I am seeing some folks struggling with their MDM 9.3 install, asking for help. What you got to do is you have to be quite careful with the steps and not miss anything. For example, check your anonymous IIS account, check IIS manager web extensions, check theDRM console for errors, check the database login and password. One o the most important thing is to ping your database or server if it doesn’t sit locally to ensure you have connectivity because this one thing account for most of the issues.

An example is you’re installing with an Oracle database on another box. The DBA has it all setup but you forgot to install the Oracle Client and you cannot connect to that database without it, even if you specify it when install DRM.

VMWare is a great technology where your one single PC can be virtualize into many, or run multiple OS, and a mix and match of these. Imagine your XP machine running Vista, Windows 2003 Server, and other hybrids like XP Media Edition without re-installing your machine from scratch. Now imagine you can run any one of these whenever you feel like needing it. It’s very simple yet powerful concept. I’s the concept of virtualization…

I have been playing with VMWare for over a year and it has helped me tremendously testing various technology. With my Quad Core system, it’s blazingly fast. Still, over time, the systems get very big — like installing Oracle Hyperion EPM 11.1.1. How often do we hear about complain on VMWare Slow Disk, VMWare Optimization, VMWare boost performance? Alot!!!

One of the most important tip I learnt is ensuring the disk is in good shape. It maybe painful but defraging your virtual OS’s hard disk is important. But the catch is you must do this in a number of steps:

1. Defrag your virtual OS hard disk using the native Disk Defragmenter under Start | Accessories | System Tools

2. Next is using VMWare’s defrag tool under VM | Settings | Hardware | Hard Disk | Utilities | Defrag

3. Finally you need to do it on the host’s machine (the real OS not virtual) Disk Defragmenter.

These can take a long time to do but if you don’t it’ll be slow and you can see it when the CPU is not doing anything and yet the machine is halting. Also do the defrag before you do a snapshot because un-defragged disk are carried forward to the next snapshot (in VMWare Workstation).

There’s a couple of good resources online and I just want to share them here with you:

• I need more performance out of my VMware environment 
• Citrix on VMware Recommendations
• VMWare Workstation Performance Tuning

Enjoy!!!

In our quest to build a financial hierarchy, in the last blog entry, we taught you how to prepare a Parent-Child hierarchy and then using the Parent-Child Import Automator tool, we created the necessary DRM Automator script without even trying. This literally saves days of work. Imagine if you’re trying to develop 10 hierarchies (with 1,000 members each)  in DRM as an example, you could literally accomplish all of it with the aid of this tool (end-to-end)  in less than an hour (or two). If not, each hierarchy could take you a day to do. That’s a big difference.

Anyway, let’s move on and load the Automator script we created into Oracle Hyperion Data Relationship Management.
Read the rest of this entry »