At COMMON I heard a couple of things.
First, at a Pete Massiello presentation it was recommended to purge
deleted records by doing a display file description to an output file and
querying that. All well and good, but you can bypass the output file and
just do this
select NUMBER_DELETED_ROWS, SYSTEM_TABLE_SCHEMA, SYSTEM_TABLE_NAME,
order by number_deleted_rows desc
However, the performance is really slow. Which brings me up to my
question: Has IBM improved the performance of these view with either TR5
(out now) or the database group (due in November)? I thought I heard that
at COMMON but I can't find any references to this anywhere. Is this true?
Has anyone done some time trials on this?