Comments on: Customers to Cloud Providers: ‘Fess Up, and Fix It

By: surj

surj — Fri, 17 Jul 2009 20:52:18 +0000

Recent cloud outages are not indication of what the future promise is.
It’s called teething trouble.

I cannot think of one technology that has NOT had teething trouble.

By: surj

surj — Fri, 17 Jul 2009 20:37:35 +0000

I agree with Derrick. Cloud won’t be adopted until enterprise grade support and SLA’s are available.

Also the great thing that IBM, Microsoft, Savvis etc do? They provide someone at the end of the phone. They have 30+ years of experience at this.

You try and get someone to talk about your problems in real time with the lower cost cloud vendors. I bet it won’t happen.

Cynically speaking – in Enterprise IT, a major part of the decision is not just the quality of the product, but who can you call and who will take the blame, when things go wrong

By: Derrick Harris

Derrick Harris — Tue, 07 Jul 2009 15:06:52 +0000

I think you’re on to something and, in fact, I would argue that users of free services shouldn’t even expect high QoS. This is why I believe successful enterprise-grade cloud providers will offer contracts and SLAs that force them to have some skin in the game. This is where Savvis, AT&T and their ilk could excel.

By: robg

robg — Mon, 06 Jul 2009 05:51:25 +0000

It’s definitely the wild, wild west again in this space, reminiscent of the ITSP boom that’s still, well, booming now. The dynamics of the situation seem to be the following:

1. Cloud providers have tiptoed into the water like the first day the pool’s open… slow, scared, hoping nobody’s watching and ready to pull back at a moment’s notice.

2. The so-called transparency may be a sign of the times in general, but I’d be willing to bet that it’s more a move to reinforce the ephemeral, laboratory-like nature that they wish (need) to convey about cloud-based architecture. Basically, by blogging, twittering, open Kimono-ing and acting diametrically opposite of the previous paradigm of iron-clad, locked down SAS-70 architecture, this sets up an expectation that “we’re all in this together, let’s cross our fingers.”

3. Caveat Emptor / You Get What You Pay For: since the commitment and incremental costs for cloud-based services are so low, there’s no leverage from the customer. A 100% recourse SLA doesn’t have a lot of teeth, and the slice/instance/processor-hour nature of things keeps everything at such a low financial risk that the providers are probably not incentivized to act as pragmatically as the old guard did.

None of this is to say that any provider isn’t doing the best they can. This is just my observation of this wave of services. Back to my reference to the ITSPs. When every mom and pop ISP began selling SIP termination as soon as they bought a PRI, the telephony market saw a whole new paradigm of “rates first, service last” lemonade-stand business models. Fortunately, the cost of entry for selling cloud space (to more than your friends and family, that is) is slightly higher.