site_archiver@lists.apple.com Delivered-To: darwin-kernel@lists.apple.com On Mar 31, 2006, at 11:38 AM, Ochal Christophe wrote: On Mar 31, 2006, at 6:00 AM, Ochal Christophe wrote: I went to the location again today, and have some more crashlogs: PowerMac G5 with OSX 10.4.5: Fri Mar 31 10:37:18 2006 Given how off the wall this address is, I'll ask the obvious question and ask if you've run Apple Hardware Test on the disc that originally came with your machine. I wonder if you're not seeing some RAM or graphics card weirdness... On a hunch, check the wall power at the site. -- Terry Replacing the gigabit switch with another switch didn't solve the problem, that is to say, atleast one machine still crashes (the one upgraded to OSX 10.4.5). _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/tlambert%40apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.a... Op vr, 31-03-2006 te 06:31 -0700, schreef William Kucharski: Unresolved kernel trap(cpu 1): 0x300 - Data access DAR=0x000000007500021C PC=0x00000000000A4638 Latest crash info for cpu 1: Exception state (sv=0x5DF25C80) PC=0x000A4638; MSR=0x00001030; DAR=0x7500021C; DSISR=0x00200000; LR=0x000A4584; R1=0x2C5B3D80; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x00038EA0 0x00038088 0x00265048 0x00264EB4 0x002AA2E4 0x000ABEB0 0x793E4346 backtrace terminated - frame not mapped or invalid: 0xBFFFFB70 Yep, found nothing wrong on all 4 machines, they work properly on another location, it's has to be something specific for the setup of the customer, alas, i can't find what. So "the only thing diffferent", apart from the network route, is the location? The easiest way to rule this out, without an electrician and monitoring apparatus that can record voltage & current fluctuations over time is to take a cheap UPS that includes a line conditioner and put it between a machine having the problem and the wall, and see if the problem goes away for that machine. Some sites have institutional clocks that spike the line voltage once an hour to synchronize (which is why the clocks in your high school went "thunk" and jumped once an hour before the bell rang). The other machines weren't being pushed as heavilly as this machine, the client will let me know if the other 3 still crash during the course of the next few days. I can imagine one machine suddenly dying, having 4 machines exibit random kernel panics starting on the same day however, isn't my idea of fun :( It's starting to look more & more like we'll have to do 2 way debugging on them, i'm hoping someone here has some more idea's as to what might be causing the machines to go belly up. This email sent to tlambert@apple.com This email sent to site_archiver@lists.apple.com