Eureka! The problem is solved through the combined application of Drake's, Ted's and Eric's solutions.
Eric: You were correct. To my chagrin, there was in fact an NT administrator loose in the building, gnawing at the wires, and telling me lies. He told me, for example, that the three machines in question were _not_ on the same UPS, and that other non-linux machines _were_ on the same UPS as some of the rebooting machines. As I should have suspected immediately, he was lying to cover some more heinous scheme. His evil plan has been exposed, and the people are all following the most excellent path once again. Ted and Drake: You are correct. The "reboot time" shown in the logs reflects the CMOS clock, as xntpd has not yet taken over. The machines were in fact all rebooting at roughly the same time (believe it or not, one machine really does take 3-5 minutes longer to reboot than do the others). Having realized that, I decided it was time to get my knees dirty and crawl around in the server room. Whaddya know? They were all on the same [faulty] UPS. I'm such an idiot. Many thanks to y'all for your help and patience. Autumn P.S. The intern has accused me of pessimism. He feels that I should call it an IPS, rather than a faulty UPS. > Are the machines rebooting simultaneously (i.e. at the same absolute clock > times, allowing for system clocks differing), or at quite different absloute > times (but with matching inter-reboot intervals)? > > If the former, consider your power supply. > > If the latter, then the Society for the Investigation of Paranormal Phenomena > would like to hear from you. > > Best wishes, > Ted. > > -------------------------------------------------------------------- > E-Mail: (Ted Harding) <[EMAIL PROTECTED]> > Date: 16-Jun-98 Time: 19:58:03 > -------------------------------------------------------------------- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
