Hi Dave, Seasons greetings to you. Wow an RVA, now theres a blast from the past, but dont worry I remember this box from 1992 when it was an Iceberg
The complete set of manuals for an RVA of this time period is: * IBM RAMAC Virtual Array Storage Planning, Implementation and Usage Guide, GC26-7170 * IBM RAMAC Virtual Array Storage Physical Planning Guide, GC26-7169 * IBM RAMAC Virtual Array Storage General Information, GC26-7167 * IBM RAMAC Virtual Array Storage Introduction, GC26-7168 * IBM RAMAC Virtual Array Storage Operation and Recovery, GC26-7171 I had a quick look at the IBM web site and couldnt easily find them. You might want to look at this sites archives and just look for T82: http://bama.ua.edu/cgi-bin/wa?A1=ind0011&L=ibm-main Best place you can look is http://www.redbooks.ibm.com/redbooks/pdfs/sg244951.pdf, which has a plethora of references for the T82. >From an I/O viewpoint the manual states: Up to four 3990 Model 3 Storage Controllers can be defined, each with up to 64 3390 and/or 3380 volumes, for a total of 256 devices. An eight-path RVA subsystem can process eight concurrent data transfer operations, and an additional eight I/O operations. This would require that 16 host channels are attached. It is less likely that an RVA T82 will be channel constrained, and customers may be more likely to attach eight channels only to the subsystem. Additional channels will provide additional channel processing capacity, but it is most likely that customers will consider a maximum of 10 channel connections to a single RVA T82. IOCDS, or the HCDGEN, must have a logical control unit (LCU) defined for each group of 64 functional devices. Each LCU should have two CNTLUNIT macros, one for each cluster. See RVA IOCDS Definition Example on page 438, and the RVA Planning, Implementation and Usage Guide for further information about IOCDS for the RVA. So, based on the above, maybe you need to give a little thought as to your physical versus logical configuration. If I remember rightly, but this was a long time ago, RVA 2 was all about Turbo and performance gains via SnapShot and throughput, bit from a subsystem viewpoint with faster HDAs and bigger caches and physical connections. ---------------------------------------------------------------------- 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
