We have 8 paths to each chpid on the mainframe dasd. And I/O subsytem. SAN is network and is fiber. I had wondered if the paths had helped swing the stats.
Richard Troth wrote: > On Fri, 7 May 2004, Ann Smith wrote: > > The performance on the mainframe dasd (STK raid 5, turbo) > > was better than the SAN. > > Interesting. Even encouraging (from the perspective of a > long-time enthusiast of the mainframe I/O subsystem). But why? > > SAN is a network, so there could be network congestion. > But since the introduction of ESCON, so is the IBM channel (a network). > Also, the STK raid 5 turbo may outperform [whatever you used] SAN, > but that is one comparison. How well would EMC, IBM, or others perform > in the same comparison? And does it illustrate a difference between > SAN and channel or between vendor and vendor? I'm also curious about > ECKD emulation overhead as an advocate of fixed-block storage. > > -- R; > > ---------------------------------------------------------------------- > For LINUX-390 subscribe / signoff / archive access instructions, > send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit > http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390