> Selecting other indexed values in the same file which gives me 5 records
out
> of the 5 million is very very fast.

30 second certainly sounds excessive but an index entry leading to over
500,000 records is not a good idea. To be effective, an index needs to lead
you to a very small proportion of the data. The bigger this proportion is,
the less benefit you will get. With an eleven character record key (assuming
your template reflects the real structure), the index record for this
customer is over 6Mb long. Everytime you update a record for this customer,
Unidata has to read a 6Mb record, update it, remove the old entry and add
the new one.

Given the number of records that you say are in the file, you are probably
going to visit most groups anyway so the index is not gaining you much for
this example.


Martin Phillips
Ladybridge Systems
17b Coldstream Lane, Hardingstone, Northampton NN4 6DB
+44-(0)1604-709200
-------
u2-users mailing list
[email protected]
To unsubscribe please visit http://listserver.u2ug.org/

Reply via email to