Well, nuttier anyway. Oracle 9.2.0.7 on AIX5.3L. IBM DS4800 SAN with CIO mounted RAID5 Oracle Datafile filesystems. init.ora filesystems_io parameter can be "ASYNCH" or "SETALL" - it doesn't make a difference. TSM5.5 for tape backup. So here's what happens - RMAN parameters - maxopenfiles=1, maxsetsize=250G, device type sbt_tape parallelism=2 (there are 5 drives). Full backup: System runs 'normally' (whatever that means), people can work, jobs complete within expected time limits, etc. Statspack snapshot covering period when backup is running shows top 5 timed events as follows: Top 5 Timed Events ~~~~~~~~~~~~~~~~~~ % Total Event Waits Time (s) Ela Time -------------------------------------------- ------------ ----------- -------- sbtwrite2 20,616,329 46,382 62.39 CPU time 9,302 12.51 db file sequential read 1,086,174 9,296 12.51 async disk IO 10,541,867 7,678 10.33 db file scattered read 145,084 1,282 1.72 ------------------------------------------------------------- Incremental backup (Level 1): System slows to a crawl. People complain. Jobs take hours longer than expected (it's like they don't even start until the backup completes). Statspack snapshot covering period when backup is running shows top 5 timed events as follows: Top 5 Timed Events ~~~~~~~~~~~~~~~~~~ % Total Event Waits Time (s) Ela Time -------------------------------------------- ------------ ----------- -------- async disk IO 9,805,203 45,977 59.34 db file sequential read 2,401,975 23,764 30.67 db file scattered read 130,897 3,458 4.46 CPU time 2,790 3.60 sbtwrite2 314,812 895 1.15 ------------------------------------------------------------- Didn't have this problem when we were using plain old jfs2 filesystems mounted without cio. What is the difference between disk access for the two types of backups?