My comment about PDSEs was just a question about using Linear VSAM data structures with load modules as the module disk image "could" be very fragmented, whereas in a PDS it would be on consecutive tracks. Although IBM has tuned VSAM and Linear VSAM datasets for performance. The underlying physical records for LV are 4K blocks which require longer CCW chains for the same amount of data considering that we normally use 1/2 track allocations for PDS loadlibs. ( Unless you're IBM and still use 6144 or 3120 or some other track wasting value as a default on distributions. )
---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
