In a message dated 12/18/2007 7:58:36 A.M. Central Standard Time, [EMAIL PROTECTED] writes: >we have a mainly "batch" workload, and the main benefit of PAV's is apparently on multiple reads. Not true. The main benefit is to reduce IOS queueing. If your batch workload is being impacted by large IOS queue times on device X, then you might want to try creating at least one PAV alias for device X. Check RMF or other I/O monitors that report on IOS queue time (e.g., TMON/MVS). IOS queue time builds up when there is more than one I/O request at the same time to the same device regardless of whether the I/Os involved are read-only or not. If you have two or more jobs running concurrently that access the same or different data sets on device X, regardless of whether the applications are reading or writing, you will accumulate IOS queue time for that device if you have no PAV aliases for it. If you have one job that accesses two or more data sets concurrently that are all on device X, you can also get queue time. The cause is multiple I/Os to the same device at the same time from whatever source, regardless of what data set on that device, and regardless of read/write. You might also be impacted by I/Os from another LPAR if those I/Os are writing into the data set from which you are reading on your LPAR. Bill Fairchild Franklin, TN
**************************************See AOL's top rated recipes (http://food.aol.com/top-rated-recipes?NCID=aoltop00030000000004) ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
