Comparing ETL tools - SSIS and Informatica

It's really a very hot topic of debate, as it's hard to find a compiled material that provide an in depth comparison on this. Would be greatful to anyone who can add a link to this topic via comment.

Thanks for the effort of Conchango in publishing a document (available for download) that dates back to 2005 (Special thanks to Jamie Thomson) . Sure that you wont be convinced with the answers of experts on thread raised before on this topic, but the assessment catalog is very useful indeed. You'll be impressed on pricing provided by Microsoft when you look at the Cost comparison made by Vincent.

Techweb Readers' Choice Awards votes SSIS as a best ETL software. Though I am a great fan of Microsoft Sql Server, I was convinced by Vincent's recent post that Informatica and Oracle are still the leaders.

Proud to see that SSIS stands atleast among the top three in ETL Survey 2007 .

I've eclectically tried to compile a standard set of criteria based on which evaluation or rating can be made:

1) Platform Support
2) Product Cost
3) Training Cost
4) Metadata Architecture
5) Connectivity
6) Reusability
7) Metadata Management

• Ability to capture metadata
• Ability to provide comments and documentation on meta objects/Self Documenting
• Ability to reverse-engineer source meta data.
• Ability to capture business metadata
• Data lineage Reporting capabilities
• Ability to report Metadata over the web
• Integration of metadata with BI Tools/Modeling tool integration/Metadata sharing with other products
• Metadata dictionary availability and accessibility
• Metadata exportable in XML format/Stored in open format for sharing
• Ability to provide target and processing impact analysis for meta data changes
• Extensibility

8) Ease of Installation, Configuration and Deployment
9) Method of Loading Multiple flat files
10) Setting the destination at runtime
11) Performance

• Minimize network traffic
• Performance Monitoring and Tuning Tool
  Optimize SQL - (eg: db indexes, etc.)
• Parallel support (load balancing)

12) ETL Building blocks

• Package Configuration
• Staging of Data from files
• Data cleansing Techniques
• Look ups
• Slowly Changing Dimension Support
• Surrogate Key Generation
• Aggregation
• Security
• Design time debugging features

13) Enterprise ETL / EAI functionality

• Support for modular development
• Integration with common packaged applications
• Web Services support
• Message queue middleware support
• Collaborative development