DSXchange

Revisiting this question...

Running on IA 11.3.1.2
Data source: SQL Server

We are trying to access an nvarchar field which the data owner advises could be up to 2GB. IA says it 'Cannot Analyze' the data. Every rule that references it fails. I've even tried LEFT(fieldname, 250) but again the mere reference to the field makes the job fail.

I've checked with our technology support and they assure me nvarchar support is already turned on in odbc.ini.

Anything else we could look into?

you should make a new topic rather than revive 5+ year old threads.

A lot may have changed since then in patches etc, regardless of version #s.

Not sure about the issue, but you will have more success on a new, unresolved topic.

And... done. Split from this topic.

See if either of this is related.

http://www-01.ibm.com/support/docview.w ... wg21376824

http://www-01.ibm.com/support/docview.w ... wg21698836

As soon as the metadata specifies more than 500 characters, IA will not attempt to analyze. The reasoning is that meaningful statistics are unlikely to be found in such fields, which are typically comments, descriptions and the like - text, basically.

What would you expect out of such a field? What would the frequency distribution look like? Even if you found 100% cardinality, there's no database in existence that could or would accept such a data type as a key column; imagine the size of the primary key index!!!

Thanks for all the input. Our IT support are exploring updating the drivers. So far, no luck. I'll report any progress/fix.