John McKown wrote: >What would you like, specifically?
I am trying to understand how to handle the large distribution sizes with the small buckets I am given to use. My users, being devlopers, can and will use everything they can find on the distribution. Yet I have to manage large DASD subsystems where subchannel I/O needs to be optimized to get the best performance overall and large volumes w/o PAV's are a bottleneck on shark (or RVA for that matter). >1) Create a 3390-9 or -27 for the /usr partition. The >problems are (a) >possible I/O queueing to the physical device, (b) not >have a >correspondingly >sized volume for D.R. purposes. The D.R. is a good point - the D.R. site would have to have the same exact hardware (-9 or -27 volumes) which would certainly limit D.R. to shark. Mark's point is much the same - recovery using LVM or RAID can be a problem. I think John Summerfield also mentioned this in a different thread. Option 2 (/usr subdirectories on different volumes) is operationally probably the best. You're not bound by a particular physical DASD subsystem and you have direct access to the volume for fsck and other repair actions. But, unfortunately, it seems you have to use RAID or LVM during your initial install to find out how big the subdirectories really are. ===== Jim Sibley Implementor of Linux on zSeries in the beautiful Silicon Valley "Computer are useless.They can only give answers." Pablo Picasso __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com
