... >We have another IDENTICAL system (same hardware, same OS level, same TSM >client level, etc) that has 3-5 MILLION MORE files and it backups up, >daily, just fine ! > >In fact, until this problem started occurring, the other IDENTICAL system >would backup in 4-6 HOURS LESS than this system eventhough the amount of >files being inspected and backed-up (including GB transferred/backed-up) >on this system is LESS. ...
Okay, then... You have there an ideal situation for tracking down the problem. What comparative analysis has been done so far, such as FTP testing for network speed, MTU sizes, actual ethernet speeds, paging rates, quantity and location of paging space within disk areas, positioning of AIX jfslog (ideally, in the middle of a logical volume, or on a separate volume in special circumstances), relative rate of packets hitting the two systems (packets generate interrupts, and interrupts are the bane of system performance), packet routing (going through a router adds time), etc? The truth is in there. Of course, that no two systems can be truly identical. File system fragmentation alone makes systems perform differently; and there have to be differences in workload, file complements, file contents, active processes, and the like. back outside to do more painting, Richard Sims