on Fri, Feb 16, 2001 at 08:27:55AM -0800, Nate Amsden (aphro@portal.aphroland.org) wrote: > SamBozo Debian User wrote: > > > > Hello to the group, > > Recently there were comments made as to the "foolishness" of > > rebooting just to reset an edited config file. How about a list of > > the cli entrys > > try avoid rebooting whenever possible. i had a bad experience with > rebooting not too long ago. a sun ultra 10..up for about 130 > days..shut it down to move a UPS, it never came back up. spent the > next 15-20 hours rebuilding it. This isn't a fault of rebooting, it's a latent hardware issue that was exposed by rebooting. While uptimes are great, there's no need to senselessly pursue uptime records. PG&E locally pretty much assures I can't, even with a UPS. Running through a boot sequence after major system reconfigurations, or just on a periodic or scheduled basis can be _very_ good practice -- it allows you to identify issues at a time you're prepared to deal with them, rather than when the problem finds it convenient (and you, most assuredly, will not). GNU/Linux provides the luxury of letting you choose your own reboot frequency, and on a stable, unmodified box, uptimes of months or years are possible. One of my favorite anecdotes along this line: after one phone switch failure -- not sure if it was the Chicago area fire, or the rolling failure that hit in the late 1980s or early 90s, AT&T was faced with the prospect of having to restart a system _that had never been shut down since it went operational_. The switch had been upgraded -- SW _and_ HW -- in place, over the years. Provide service guarantees through redundancy, not through pushing HW past its limits. -- Karsten M. Self <kmself@ix.netcom.com> http://kmself.home.netcom.com/ What part of "Gestalt" don't you understand? There is no K5 cabal http://gestalt-system.sourceforge.net/ http://www.kuro5hin.org
Attachment:
pgpRDWUB2S2yy.pgp
Description: PGP signature