On Thu, 20 Apr 2006, Alan DuBoff wrote: > On Thursday 20 April 2006 06:19 am, Dennis Clarke wrote: > > Some people are soooo stuck in thinking "it still works and so we are not > > going to fix it" attitude. When I see yet another E4000 ( not E4500 ) > > running Oracle for 70 people I just want to scream. Really. Running > > Solaris 8 of course. Its maddening when I _know_ they can get a X2100 with > > a pile of RAM and support and totally replace the box. The electricity > > savings alone will justify the increase in throughput. > > > > Drives me mad. Or at least more mad. > > Understood, but it's at least worth pondering how Sun's customers running > these legacy systems will move forward. Eventually they'll need to upgrade. > Will they replace with new SPARC systems, or will they replace with new > systems of another architecture. Solaris now offers them more options, and > other architectures offer more options other than Solaris. Wether these folks > keep these systems on Solaris or not will be determined. My guess is that > commodity hardware running Solaris will be the most attractive for those > folks.
It has nothing to do with "it works, so why fix it?" The reality is that (sane, anyway ;-) customers will replace systems (on their normal replacement cycle; quicker if cost savings or performance needs justify) gear with whatever is supported by the ISV that supplied the app that sits on them. I have racks of boxes just like the ones Dennis describes -- old Ultrasparc II / III gear, Solaris 8. I'd love to move those systems to Sun's AMD64 hardware and Solaris 10. I'd even love to move them to Solaris 10 on Sparc. I can't do either if I want application support I almost wish Sun would go ahead and EOL Solaris 8 to force a little movement on the ISV front ;-). That, and Oracle not double-licensing for UltraIVs, are my pipe dreams for the year later, chris _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
