On Sat, Feb 25, 2012 at 6:13 PM, Simón Ruiz <[email protected]> wrote: > On Sat, Feb 25, 2012 at 11:59 AM, Brent Foor <[email protected]> wrote: >> Had this issue when some ram went bad try poping out a stick at a time and >> booting it. > > Have recently had the *strangest* experiences with RAM. On computers > in our computer lab which worked fine for *years* with the RAM they > had, a few randomly started not booting. > > They came with 2 512MB sticks, we added 1 1GB sticks when they wanted > to use the lab for PhotoShop and it complained. > > Memtest86+ reported no RAM errors. > > Figured out finally that if we used only one kind (brand/size) of RAM > or another it would work, but not with both. > > That is, each individual RAM chip was fine and usable, but suddenly > not together. > > Later, a co-worker brought in her computer which was not booting, and > I took a look at the RAM. > > If I left stick A, but not B in it would work. B but not A would also > work. Either stick with a foreign stick C would also work. But no > combination of A with B would work. > > Same brand, same capacity chips. Again, Memtest86+ reported no RAM troubles. > > RAM can definitely be a fickle, fickle beast. > > You might, (depending on how many discrete chips you have) rather than > unplug one RAM chip at a time (or after that process, if it doesn't > yield results) unplug all *but* one RAM chip at a time. Just a > thought. > > I've got similar weird stories about swapping around power supplies, > earlier this year, to confirm Nestor's advice. > > For all I know, this behavior might have been a reflection of a power > supply or motherboard issue that was somehow resolved by swapping out > RAM (different RAM chips need/use/tolerate/affect voltage > differently), but the problem was resolved, so I had other more > pressing matters to worry about that coming up with an academically > sound diagnosis. > > Just some of the multitudes of examples of how theory and practice differ. > > Especially when it comes to computing: In theory, theory and practice > are the same; in practice, they are not. > > Hope you're all doing well. > > Simón
Just to support everyone's comments and point to Indiana as part of the problem. In the past I have reviewed systems in large commercial installations that have showed like results. We would phase in the secondary systems and then test the bad systems. In one case an not to be named data center had zinc particles floating in the air that were attracted to the components. In one case a none backed up produciton system was restored by plugging in the IDE cable (it had wiggled loose after years of hot/humid cool/dry weather). And Simón is 1000% spot on. Remove all but the base needed to start the system. Just like science class, you remove all variables for testing. Unplug the HDDs, CDROMs, printers and any thing connected. In Indiana weather re-seating the components can have a dramatic effect if the copper or gold needs a fresh scratch. -- ~ Andrew "lathama" Latham [email protected] http://lathama.net ~ _______________________________________________ Fwlug mailing list [email protected] http://mail.fortwaynelug.org/mailman/listinfo/fwlug_fortwaynelug.org This is a public list and all posts are archived publicly. Please keep this in mind before posting.
