[nycbug-talk] FBSD sysutils/bsdstats

[Ray Lai]

On Sun, Aug 13, 2006 at 10:13:42PM -0400, Dan Langille wrote:
> On 13 Aug 2006 at 20:42, Ray Lai wrote:
> 
> > On Sun, Aug 13, 2006 at 07:56:05PM -0400, Max Gribov wrote:
> > > Brad Schonhorst wrote:
> > > 
> > > >On 8/10/06, Jeff Quast <af.dingo at gmail.com> wrote:
> > > >  
> > > >
> > > >>On 8/9/06, michael <lists at genoverly.net> wrote:
> > > >>    
> > > >>
> > > >>>Funny.. the OP in those links put up another link as a reference:
> > > >>>(http://www.mreriksson.net/uptimes/) The Uptime Project.  I never
> > > >>>really understood that cheesy bravado.  Why would you want to brag that
> > > >>>you have not patched your box for the last 6 years?  And not just brag..
> > > >>>but to register it and have it available on a website, along side
> > > >>>other poorly maintained boxen.. like it is a badge of honor.
> > > >>>      
> > > >>>
> > > well, you only have to reboot when you patch the kernel.
> > > fbsd at least has a pretty good record of not having too broken of a
> > > kernel, and sometimes there are issues which one does not have to patch
> > > since they were removed from the custom compiled kernel.
> > > 
> > > uptime can be considered "badge of honor"  in a sense that the os you're
> > > using does not have too many kernel issues, the fact that you made a
> > > custom enough kernel not to worry about useless issues, and plus the
> > > fact that things rarely go wrong in the bsd server world.
> > 
> > I generally reboot machines after daemons are patched as well, in case
> > the startup process got trashed due to config changes or something.
> > It's better to find out immediately rather than after you have forgotten
> > the change six months later.
> 
> Are you speaking theoretically or do you have practical examples of 
> such instances?

Of course it's happened to me.  Newer daemons are sometimes missing
libraries or no longer accept old config file syntaxes.

-Ray-