Keep your UCM instance in shape

inshape_small.jpg
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.

idcanalyzer.jpg

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.

Comments:

I wanted to add that admins should keep their eye on the "Steps to Maintain - UCM Content Server - Oracle Universal Content Management 10.1.3 (10gR3)" metalink article. Oracle publishes these articles which include all patching information. It's a great way to keep on top of all the patch releases. It's the extra, above-and-beyond steps like running the content analyzer and ensuring your instance(s) are all patched up that will really help reduce stress.

Posted by Ryan on November 06, 2009 at 08:29 AM CST #

I wanted to add that the management of our production environment is not the primary concern at the moment. We have a large number of paralell projects running at the same time. These projects each have their own environment that they are developed on. Metadata changes are made there and propagated to a merged environment. The lack of tools to compare metadata models and values therein makes this a hard work. I am currently trying to write a webapp using CIS that does this and its a lot of work. Any tools you know of thet could help? Thanks for a great Blog! /T

Posted by Tony D on January 04, 2010 at 12:15 AM CST #

Hey Terry,
Yeah, I don't know any built-in or custom tools to do metadata comparisons. The only thing I can think of that is close are the Active Reports available through the Library. You can define a report against 'Information Fields' to get a report on the configuration of custom metadata fields. But this report is fairly light and doesn't cover things like dependent choice lists or schemas/views. If you wanted something that can provide a more complete view and diff between metadata models, writing something custom in CIS/RIDC is probably the way to go.

Posted by kyle.hatlestad on January 04, 2010 at 01:26 AM CST #

Hi Kyle, We are implementing a very large URM project, and are looking to set the idclogs to roll by size rather than just by day. Is there an easy way to manage the logging? Any of those nice secret config entries? Thanks! Terry

Posted by Terry Y on May 13, 2011 at 08:17 AM CDT #

Hey Terry, I'm afraid I don't know of any way to change the behavior of the logging. I believe it can only roll based on days and not size. Thanks, Kyle

Posted by Kyle Hatlestad on May 23, 2011 at 01:33 AM CDT #

any idea of why im receiving a lot of this messages when i run idcanalyze?

----------------
idocscript08.08 22:58:25.675Thread-13Script stack dump (user: <undefined>, request: <no path>).
No resource include or function stack for error.
Evaluated value: <$ActiveIndex$>. Offset in runtime evaluation is at character 3. Could not find the parameter 'ActiveIndex'.
----------------
idocscript08.08 22:58:27.616Thread-13Script stack dump (user: <undefined>, request: <no path>).
No resource include or function stack for error.
Evaluated value: <$ActiveIndex$>. Offset in runtime evaluation is at character 3. Could not find the parameter 'ActiveIndex'.
----------------
idocscript08.08 22:58:29.286Thread-13Script stack dump (user: <undefined>, request: <no path>).
No resource include or function stack for error.
Evaluated value: <$ActiveIndex$>. Offset in runtime evaluation is at character 3. Could not find the parameter 'ActiveIndex'.

Posted by guest on August 09, 2012 at 03:04 PM CDT #

Hi Kyle,
Is there any way in ECM 11g,that while checkin the document it should be
link/store in Content Folio or we need to select available folio for checkin the documents.

Kindly suggest any solution or alternative for the same.

Posted by Prashant on March 04, 2013 at 07:07 AM CST #

Hello Prashant,

Yes, you must start from the Content Folio and choose to check-in from there in order to associate the document to that folio.

Thanks,
-Kyle

Posted by Kyle Hatlestad on March 06, 2013 at 03:38 PM CST #

Content Server Analyzer login failure. The username and password is working fine on content server console, but not on Analyzer? Please advise?

Posted by guest on May 01, 2013 at 09:13 PM CDT #

Post a Comment:
  • HTML Syntax: NOT allowed
About

Kyle Hatlestad is a Solution Architect in the WebCenter Architecture group (A-Team) who works with WebCenter Content and other products in the WebCenter & Fusion Middleware portfolios. The WebCenter A-Team blog can be found at: https://blogs.oracle.com/ ateam_webcenter/

Search

Archives
« April 2014
SunMonTueWedThuFriSat
  
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
   
       
Today