Re: [SLUG] DNS BIND: what's happen....?

[Ben Leslie <benno@xxxxxxxxxxxx>]

On Thu, 02 Aug 2001, Jamie Wilkinson wrote:

> This one time, at band camp, rod@xxxxxxxxxxx wrote:
> >The orginal poster was making the perfectly valid point (probably 
> >based on experience) that its worth rebooting after configuration 
> >changes to ensure that the system will behave as required next time 
> >it reboots.
> 
> I so disagree with you.
> 
> The only time you *should have to* reboot is to add hardware or change
> kernels.
> 
> >Now what happens when 2 of the 30 changes made to that machine 
> >over its last 300+ days of uptime are gone because they were'nt 
> >added to rc.xyz correctly after they were done by hand on the 
> >console?
> 
> Okay, you have a point.  I'm going to blow my own horn and am probably going
> to be a bit hypocritical here -- I always doublecheck and test, even halting
> and restarting services to make sure they come up right, especially starting
> them as the bootscripts would.  On d*b**n systems, this is as simple as
> [WARNING: Content removed due to possible distribution advocacy]

I have to agree, I don't think rebooting the machine just to overcome
sloppy admin is the solution. It is probably fine and easiest for a home
machine but for a production server it isn't really the answer. Having some
sort of routine, which includes documenting the change and the ensuring
it will start on boot if necesary, can remove the need to reboot after each 
change.


My 2c

Benno