[31936] in North American Network Operators' Group
Re: InterNAP?
daemon@ATHENA.MIT.EDU (Stephen Sprunk)
Tue Oct 31 13:37:52 2000
Message-ID: <024501c04369$67c71020$30132ca1@glock>
From: "Stephen Sprunk" <ssprunk@cisco.com>
To: "Spolidoro, Guilherme" <guilherme.spolidoro@unisys.com>,
<nanog@merit.edu>
Date: Tue, 31 Oct 2000 12:32:07 -0600
MIME-Version: 1.0
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Errors-To: owner-nanog-outgoing@merit.edu
Thus spake "Spolidoro, Guilherme" <guilherme.spolidoro@unisys.com>
>
> Since they don't own the backbone, their SLAs apply only to their
PNAPs,
> i.e. they cannot guarantee performance, packet loss, delay, outages,
etc
> over somebody else's backbone (e.g. Sprint, UUNet, AT&T, etc).
Nor can anyone who peers with the above; InterNAP theoretically has a
better chance of getting good service, since they're a paying customer.
That's how their business model goes, at least.
> Many large enterprises are leveraging on the Internet to deploy VPN
> connectivity between company sites as a replacement for Frame Relay
> or ATM PVCs. In such cases, it's very important to make sure that the
> service comes with the right SLAs.
SLAs are rarely worth the paper they're printed on. If you do not
design your network/services/whatever around the *expectation* that your
carrier will fail, you're in for a rough ride. What consolation is a
prorated refund when your circuit is down for 2+ days?
S
| | Stephen Sprunk, K5SSS, CCIE #3723
:|: :|: Network Design Consultant, GSOLE
:|||: :|||: New office: RCDN2 in Richardson, TX
.:|||||||:..:|||||||:. Email: ssprunk@cisco.com