Business case for switching to Linux

[Christopher Browne]

On 4/10/06, John Van Ostrand <john-Da48MpWaEp0CzWx7n4ubxQ at public.gmane.org> wrote:
> On Mon, 2006-04-10 at 18:52 -0700, Christopher Browne wrote:
> > On 4/10/06, John Van Ostrand <john-Da48MpWaEp0CzWx7n4ubxQ at public.gmane.org> wrote:
> > > It really depends on the system and the administrator. A re-install of a
> > > complex system that would take days to reconfigure may be more of a
> > > business hit than the hour to fix it.
> >
> > The only reason for it to be "acceptable" to get stuck with this is if
> > you have a totally unmaintainable system that is actually impossible
> > to reinstall.
>
> Not impossible, just time consuming.

It can be effectively impossible if the work cannot be accomplished
within acceptable time constraints.

Suppose it would require a couple weeks of redevelopment effort to get
a bunch of CGIs modified to run on more modern hardware.

> > If we had such a problem with a database server, you can bet it'll be
> > *toast*, down to reimaging from scratch.  It's not that hard.
>
> But if an image doesn't exist?

For Linux, an image is only a CD-ROM away.

For our AIX servers, this is the sort of thing we keep images for...

> > And what you really want with your systems is the ability to rebuild
> > from scratch, *easily*.  See <http://www.infrastructures.org/>; the
> > folks in that organization find it unacceptable if they can't rebuild
> > a server *from scratch* in 20 minutes.
>
> I can see how, in a very locked down environment, that one could recover
> from an image in 20 minutes but a re-install would take far longer than
> that.
>
> It would be nice if all servers could be restored in 20 minutes. In most
> cases it's just not possible. Even then, a data restore could take hours
> beyond the imaging time. I've dealt with customers that don't have a
> backup or an image. To re-install would take a full-backup, re-install,
> re-configure, restore, then the ensuing issues. On simple servers with
> lots of data this can take many hours. On complex servers with previous
> admins long gone, and poor documentation its a days-long event.

Sheer disaster, particularly in the case where the server was
compromised, and you can NOT trust it anymore.

If you actually need high availability...

- Inadequate procedures means you don't got it.
- Inadequate documentation means you don't got it.
- Inability to reimage quickly means you don't got it.

> > If anything, Linux should be easier than the "traditional Unixes" to
> > do this with.
--
http://www3.sympatico.ca/cbbrowne/linux.html
"The true  measure of a  man is how he treats  someone who can  do him
absolutely no good." -- Samuel Johnson, lexicographer (1709-1784)
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml