We have an IMS segment which is being read to get all records with a value matching a field. This is usually a small number.
This ended up being slow due to DVM needing to read the entire segment to "find all matches"...as I understand it, under this setup, DVM has to read the entire segment in order to "find all" since this is not a keyed field.
We proceeded to get a secondary index set up, so now this is a keyed field.
At this point, do we need a virtual table created using the secondary index or is it a virtual view? As it turns out we need to use field A to get all the matches and only need one other field returned.
Please advise...
------------------------------
Kevin Flanagan
General Motors
------------------------------
#Db2Toolsforz/OS