We're running IDS 9.4.FC5 on AIX 5.2. I have a table that is fragmented
by round robin in two dbspaces. It does have rowids (not my decision),
and a number of indices. I've been noticing that the onstat -g ppf
output for the two partitions have wildly different numbers of
sequential scans. Today, for instance, one partition has 11,000 or so,
the other has 2. How does that make any sense? Surely in a round robin
environment if you scan one side you will also have to scan the other.
This table has a pretty large number of records, so I'd like to figure
out what this means. Any ideas? Thanks.
DC