[140089] in North American Network Operators' Group
Re: Amazon diagnosis
daemon@ATHENA.MIT.EDU (Jay Ashworth)
Sun May 1 14:12:56 2011
Date: Sun, 1 May 2011 14:10:40 -0400 (EDT)
From: Jay Ashworth <jra@baylink.com>
To: NANOG <nanog@nanog.org>
In-Reply-To: <4DBDA17C.2040303@tiedyenetworks.com>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
----- Original Message -----
> From: "Mike" <mike-nanog@tiedyenetworks.com>
> On 04/29/2011 12:35 PM, Joly MacFie wrote:
> > *http://aws.amazon.com/message/65648/*<http://aws.amazon.com/message/65648/>
>
> So, in a nut shell, Amazon had a single point of failure which touched
> off this entire incident.
>
> I am still waiting for proof that single points of failure can
> realistically be completely eliminated from any moderately complicated
> network environment / application. So far, I think murphy is still
> winning on this one.
Well, in fairness to Amazon, let's ask this: did the failure occur *behind
a component interface they advertise as Reliable*? Either way, was it possible
for a single customer to avoid that possible failure, and at what cost in
expansion of scope and money?
Cheers,
-- jra
