 |
 |
Archives of the TeradataForum
Message Posted: Mon, 26 Jan 2004 @ 18:56:47 GMT
| Subj: | | Re: Misunderstanding of join index |
| |
| From: | | Victor Sokovin |
Glen,
| | NCR's recommendation has been to drop the join index before you archive and restore the join index after the backup is completed. Since
some of our join indices are non-trivial, this is not an acceptable answer. | |
Pity this limitation had not been specified and documented by NCR in the same way as they ahve done with other bulk utilities
(MultiLoad). I believe you have written about ARC and JI issues at least half a year ago and I am surprised the situation is still the same.
In my opinion, however, this additional ARC trouble does not change significantly the size of the target group of customers who might consider
using JI. The applications must already be well tuned and fast enough to be able to rebuild the JIs after applying a bulk utility, be it MultiLoad
or ARC. Or large DML statements which run faster after the indexes have been dropped.
Regards,
Victor
| |
