[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
Gordon Henderson wrote: > > Unlike Simon, I have had servers segfault due to hardware problems - > mostly memory. I'm misunderstood, I said segfault is usually a software problem. I've seen it due to hardware, but not usually on more than one server at the same time. And usually it is accompanied by huge spew to system logs, or obvious file system corruption or disk read errors (the kernel log is worth checking). The HP-UX documentation had a telling discussion on the error "Bus Error" (a similar but slightly different type of fault), which after a long list of possible software causes noted at the end that a "Bus error may also be caused by an error on the bus". Segfault just means that memory is being accessed in a way that isn't allowed, reading memory not allocated to a program, writing to read only memory. Sam really needs to be more specific - do "ls" and "vi" fail on both servers? Do they produce any other output? What Distro, what version, what changed, any error messages in logs. What happens on reboot (he said they won't reboot?). Even where I've had DELL firmware bugs cause sections memory to be randomly overwritten (and DELLs memory utilities failed to find it - memtest86+ found it), segfault wasn't the most obvious symptom (although some accesses to the corrupted file system would cause some programs to segfault). -- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/linux_adm/list-faq.html