Path: utzoo!attcan!uunet!cs.utexas.edu!mailrus!ames!amdahl!pacbell!lamc!wet!epsilon From: epsilon@wet.UUCP (Eric P. Scott) Newsgroups: comp.sys.next Subject: Re: NeXTs on & off Summary: Don't turn it off. Message-ID: <630@wet.UUCP> Date: 1 Oct 89 23:54:05 GMT References: <1923@cunixc.cc.columbia.edu> Reply-To: epsilon@wet.UUCP (Eric P. Scott) Organization: Wetware Diversions, San Francisco Lines: 36 In article <1923@cunixc.cc.columbia.edu> brad@woof.columbia.edu (Brad Garton) writes: >What's the feeling out there about turning the cube off when idle? >I know the manual says it's a good idea to shut it down when gone >an extended period (it mentions "overnight" as an example), but I >wonder about the wisdom of shaking up them electrons every morning >and night. Just because there's no one at the keyboard doesn't mean it's not being used/useful. I'm so glad NeXT added PowerOffDisabled in 0.9. I don't know WHY there's a seductive power button on the keyboard. I really don't. It's sure not intended to be used! (Must be a marketing appeal to the BusinessWeenies who only know from PCs and can't even pronounce "crontab" let alone understand it.) Just tell them "the machine is instantly available and it has a screen saver." The only machines we ever let be turned off are totally diskless, and even this is discouraged since they're all networked. Speaking of power-off, does 1.0 recover gracefully from power failures? 0.9 sure didn't, and we had problems when a big chunk of San Francisco lost power for a few hours and our nameservers weren't UPSed. At the time we were using a NeXT as primary master and a Sun-4 as a secondary, since ancient BIND wouldn't run properly as a secondary (1) it didn't transfer WKS records (2) named "went poof" a lot. [ Hence my enthusiam for NeXT's announced support for BIND 4.8. ] This was before we got an out- of-state machine to join as a secondary. The Sun came back up just fine, but the NeXT was dead, dead, dead. in.named on the Sun patiently tried to zone-transfer from the NeXT, meanwhile ypserv was forking like crazy trying to get some action out of in.named. Needless to say, the Sun didn't handle this too gracefully. The rest of the LAN wasn't too thrilled either. -=EPS=- / SFSU