Hi All,

This is a roadmap question hoping for a response from a member of the
Google App Engine Team.

When GAE finally moves out of Beta (is January 2011 too optimistic?),
there is a general expectation that developers will be able to run
mission-critical web apps with best-of-breed scaling, redundancy and
downtime. Please correct me if I'm wrong in this assumption (and if I
am wrong then that raises other obvious questions).

When that nirvana moment arrives, I expect that Google will have fine-
tuned its SLA to reflect its confidence in the robustness of its GAE
platform and associated infrastructure.

But what about GAE developers whose apps provide "critical" services
to end-users - a CRM application would be a good example. In this
case, the developer would be expected to provide an SLA to his
customers. Obviously the two main technical considerations in framing
this SLA will be the quality of the App Software (the developer's
responsibility) and the reliability of the platform/infra-structure
(Google's responsibility).

Looking purely at the platform/infra-structure side of things, a
conservative end-user SLA might water down Google uptime guarantees
just to give a little wriggle room. But the guarantees are not the
real issue here. The difficulty lies in what recourse the end-user has
to an SLA violation. If Google fails to meet it's uptime guarantees,
e.g. due to a major outage taking a couple of days to stabilise, then
what recourse can the developer realistically expect from Goggle if he
has 10,000 angry business customers?

Of course this question is pertinent to all "hosting" companies. The
critical difference here is the level of control which the developer
has in a GAE scenario compared to, say, a dedicated server cluster.
For example, in a dedicated server scenario, you could have a worst-
case disaster-recovery plan in place to get your web app back on-line
within say 8 hours. But with GAE, that level of control is completely
relinquished to Google. And it is worsened by the fact that your app
is mixed up with zillions of other apps in the GAE soup.

I'm not expecting miracle answers here. I'm thinking more in terms of
how GAE developers can approach end-user SLAs, bearing in mind that a
watered-down SLA could be a deal breaker for end-users. I'm wondering
also if Google is thinking of offering enhanced GAE services for
developers of critical apps, offering an extra level of redundancy
using a special "backup" cloud on independent infra-structure.



-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" group.
To post to this group, send email to google-appeng...@googlegroups.com.
To unsubscribe from this group, send email to 
google-appengine+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-appengine?hl=en.

Reply via email to