Sorry, no better idea on the internal structures of the index - just the obvious:
1. Was the index created on a different version of UniData or O/S? 2. Were the same attributes indexed/were the dictionaries the same? 3. Was the alternate key length any different? A list.index from before and after might have helped. Other than that, all I can say is: Free Space! Woo Hoo! Colin >-----Original Message----- >From: Baakkonen, Rodney > >Over the weekend, I rebuild the indexes for our largest file >for claim details. This file has about 41 million records in >it. The directory for the dynamic file says it uses 35 gig of >space. So it is a pretty good sized file. > >Prior to rebuilding the indexes, they were contained in 8 >separate files >idx001 - 8. After the rebuild, the indexes were contained in 5 >parts, idx001 >- 5. So I got back more than 4 gig of space back by doing the >rebuild. We are trying to evaluate now if selects using >indexes are any faster after the rebuild. ------- u2-users mailing list [email protected] To unsubscribe please visit http://listserver.u2ug.org/
