[AusNOG] Outage that costs Millions

[Andrew Fort]

On Wed, Jun 30, 2010 at 10:02 AM, Narelle <narellec at gmail.com> wrote:

> So was it a fault in a hosted service? A network failure? A router?
> Fibre cut? Software in an application? Hardware on a web service? DNS?

Given the info posted to the list yesterday, it could have also been a
customer connecting the broadcast domain of their P2P TWE service.
But yes, who knows?

> Personally I always thought it better to post a statement of what went
> wrong in a plain English explanation with an apology.

Precisely.  Union Carbide didn't learn that, BP certainly hasn't
learnt this, and I doubt Telstra has, either.  I'm less amazed that
Google did it (only once, however), when explaining a major outage,
but for a big vendor-neutral grey company that's a turn up for the
books - even if that effort was driven entirely by their operators'
management.

If you've screwed up already, surely the goal is to minimise outrage
over being lied to.

http://www.psandman.com/articles/shield.htm


> The fact is that the rush to print 'something' is preferred to good
> journalism, and our industry still hasn't improved in its openness on
> what goes wrong when it inevitably goes wrong.

There's also the lack of side taking involved in "just reporting the
facts".  This is safe for journalists - they won't be accused of bias.
 While the internet makes this "better" (facts are easier to find,
right?), it makes it harder, too - it's easier to use Google for your
fact checking rather than picking up the phone.

-- 
Andrew Fort (afort at choqolat.org)