[SGVLUG] Well its fun to be scared sometimes.

[Martin Bähr]

On Wed, Aug 31, 2005 at 03:59:35PM -0700, Alan Horn wrote:
> Regular preventive maintenance type work including reboots can be a Good 
> Thing (tm)
> Or maybe thats folklore ?

it depends on what kind of changes you make.
i recently did a debian upgrade, which usually goes fine, well, since i
was in the room where the server was, where i am not normally, i decided
to reboot. 

it failed.

turns out that through the upgrade lilo was run, and it for some reason
did not correctly take the raid we had into account.

the fix was semitrivial (had to put a cdromdrive into the machine, boot
knoppix, manually set up the raid, and then) just rerun lilo. i have no
clue why it went wrong the first time, however, had i not rebooted while
i had the best chance to fix problems, i am sure this would have bitten
us on some friday night with nobody around to fix it until monday
evening (we are all volunteer admins)

to make a long story short, if you have changes that are critical, and
you are around in case of failure, then do use the opportunity to reboot
to verify that things still work.

what kind of change is critical, depends on your experience and judgement.

greetings, martin.
-- 
cooperative communication with sTeam      -     caudium, pike, roxen and unix
offering: programming, training and administration   -  anywhere in the world
--
pike programmer   travelling and working in europe             open-steam.org
unix system-      bahai.or.at                        iaeste.(tuwien.ac|or).at
administrator     (caudium|gotpike).org                          is.schon.org
Martin Bähr       http://www.iaeste.or.at/~mbaehr/