On Wed, 13 Sep 2006, Brent Franks wrote: > On 9/12/06, Dag Wieers <dag@xxxxxxxxxx> wrote: > > I think you can download the relevant packages online and use it for trial > > purposes without needing a sales contact. At least that was possible for a > > while (not sure if they changed their policy again), I haven't looked at > > it since I switched positions. > > > > Did you look at their download site ? > > Thanks for the follow up. Yes, after an extensive search we finally > found it. It was packaged under the IBM (Cluster Server Manager) > trial-ware. > > Once we found it, it brought up other interseting questions we haven't > fully answered yet. It seems we were able to get the RPM's installed, > but all the docs point to GPFS (And CSM for that matter) are only able > to run on IBM servers. IBM only supports it on IBM servers, but it should work fine on other systems as well. Fact is that data-integrity during hardware-failure (both SAN or server components) is very hard to verify with hardware from different vendors. But ask your IBM representative about support on other hardware. They might have had success stories with your combination of hardware. > I could understand if it was just IBM AIX or Power architecture, > however the x series would seem to imply that CSM and GPFS would run > on any Intel i386 platform. Yes, when GPFS only worked on AIX, it was easier to support the different hardware combination. But since it has been ported to Linux, the same limited support has been extended to their own hardware. It is very hard to figure out what has caused data-corruption during hardware failure in a PC world (with caching going on on different levels) unless you control the hardware and filesystem. > Am I mistaken here? Is there some sort of hardware abstraction layer > that CSM or GPFS checks prior to working in production? I doubt that is the case. It's a support issue. I doubt they won't provide support for GPFS on 'alien' hardware for installation, configuration and basic troubleshooting though. Kind regards, -- dag wieers, dag@xxxxxxxxxx, http://dag.wieers.com/ -- [all I want is a warm bed and a kind word and unlimited power] -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list
