On Wed, Jun 12, 2002 at 06:21:31PM -0400, Kevin Darcy wrote:
> 
> Big deal. You could have a cron job do "health checks" and change the A
> record via Dynamic Update.

Easier to debug, that's for sure.

> The real issue here is that short TTLs are anti-social. That's why
> responsible architects/admins don't implement such scripts. Unfortunately,
> this reticence has opened up the door to a bunch of hucksters overcharging
> for their "WSD" devices (or GSLB, or whatever the nom-de-jour happens to
> be). I was hoping the dot-bomb phenomenon would have wiped out all of those
> hucksters, but apparently even the big networking companies have jumped on
> this bandwagon, so it's unlikely to die anytime soon. Sigh...

I'm not interested in arguing this, but instead would like to know what
socially responsible architects/admins do. I've had a situation where a
whole coast was unreachable, so the redundant datacenters (with the same
provider) on the same coast were useless. 

Do you use fixed IPs with long TTLs and rely on routing protocols?
Something else? Do you just accept outages across regions?
-- 
He may look like an idiot and talk like an idiot but don't let that fool
you. He really is an idiot. - Groucho Marx


-- Attached file included as plaintext by Ecartis --

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE9B9lXWpDEZMF673kRAnlfAJ0YIujidrcKGwmNh019RqLyjJEnLACfZMWC
QPpvtFNNAOKPUtVA2HIzAgw=
=vKCM
-----END PGP SIGNATURE-----