Calculating Cost

Sun Aug 22 16:43:48 UTC 2010

On Sun, 22 Aug 2010 18:29:52 +0200, Kasper Adel said:

> How would you calculate the cost of a network outage, specifically if its
> related to a software bug or a misconfiguration.

Just your actual costs, or your costs plus refunds due on SLAs, or your costs
plus refunds after SLAs once you finish sleazing out of paying off (you
providers who make opening a ticket hard enough so most outages are fixed
before the ticket is opened know who you are), or your customer's costs, or
some other number?

Calculating the cost once you have a definition is usually pretty simple
actually.  It's nailing the jello definition to the tree that's the hard part.

> Suppose that this could have been avoided by testing in a lab before
> deployment, how  can i calculate that too?

Same as above, plus merciless mocking from anybody who finds out you could have
found out about it in your lab.

Calculating the cost of running a lab large-scale enough to detect the issue is
a different and equally hard question.  Do you need to just catch the obvious
stuff, or do you need to also catch the backplane lockup that happens when you
have features A and C, but not B or Z, enabled, a prime number of interfaces, 3
of which are down, and between 57 and 61 percent capacity on at least 7 links?
(You think I'm kidding, ask the list if that sort of stuff actually happens. ;)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 227 bytes
Desc: not available
URL: <http://mailman.nanog.org/pipermail/nanog/attachments/20100822/d5f58148/attachment.sig>