Kyle's Blog on Universal Content Management

In day-to-day operations, there isn't much you need to do in regards to maintenance of a UCM instance. You may check the Work In Progress queue to make sure things are converting and indexing fine. You may check the database and file system to make sure space and performance are OK. And then maybe take a peek at the logs to look for any suspicious errors popping up. But that's about it.

But there is actually a utility that comes with the product which helps keep things in shape called the Content Server Analyzer. It's a process that runs on the server that checks the integrity of the file system, database, and search collection. The 3 major components that make up the repository. It will look for things like:

• Confirm the accuracy of synchronization between three important content server database tables (Revisions, Documents, and DocMeta).


• Confirm that the dRevClassID and dDocName fields are consistent across all revisions of content items.


• Determine if the file system (native and web-viewable file repositories) contains any duplicate or missing files.


• Ensure the accuracy of synchronization between the search index and the file system.


• Ensure the accuracy of synchronization between the search index and the Revisions database table.


• Ensure that the file system contains all necessary files.

After the Analyzer runs, it will generate a detailed log reporting on everything it found. And not only does it create the report, it can also generate scripts to fix the issues it finds. You can even choose to have it fix things for you as it runs. I'm a bit conservative, so I prefer to simply have it create the scripts and logs so that I can perform the fixes myself.

Under normal circumstances, there shouldn't be any discrepancies in the system. But software being software, sometimes things can get out of sync. Perhaps there was a server crash, network hiccup to storage, unexpected disruption to the indexing process, etc. Even though the server is up an running and everything looks ok, there may be these minor issues in the background.

To get to the Analyzer, it's an executable available on the server. The way to start it is similar in both Windows and Unix. Go to the [ucm dir]\bin directory and you'll see the IdcAnalyze executable. It's built so that you can simply run it on a command line or launch it with a gui interface. To launch it with the interface, add the '-g' option.

Here you can turn on the various options you want checked and click Start Analysis to begin. If you turn on reporting, this will create a sub directory in the \bin\ directory and place the report along with any cleanup scripts. If you run it again, it will rename the previous log directory and create a new one.

Alternatively, you can run the IdcAnalyze process as just a command line process. This way you could automate this task on a scheduled basis in the OS. Just add '-help' or leave off any options to see the available options.

For further details on the Content Server Analyzer, check out the 'Content Server Troubleshooting Guide' starting on page 3-23.