Azure One Month Pass - No Credit Card Required

Just went live - get a whole month of Azure - no credit card required http://azurepassusa.cloudapp.net

US only.

Does anyone really think that they're qualified to write good cloud applications? I can barely write a half decent multi-threaded application. I mean if it is read only then sure, but with caching I doubt the load is very high anyway. Can you imagine the IO issue you'd be presented with when half a dozen apps are writing to the same database?

ManipUni said:

Does anyone really think that they're qualified to write good cloud applications? I can barely write a half decent multi-threaded application. I mean if it is read only then sure, but with caching I doubt the load is very high anyway. Can you imagine the IO issue you'd be presented with when half a dozen apps are writing to the same database?

You don't need to worry about that.

ManipUni said:

Does anyone really think that they're qualified to write good cloud applications? I can barely write a half decent multi-threaded application. I mean if it is read only then sure, but with caching I doubt the load is very high anyway. Can you imagine the IO issue you'd be presented with when half a dozen apps are writing to the same database?

The whole point of Azure is that it abstracts all that away from you.  You just write a plain old webapp; the Azure infrastructure takes care of all the hard parts:  data distribution/synchronization, automatic allocation of nodes, etc.

blowdart said:

CannotResolveSymbol said:

*snip*

That doesn't magically remove the need to index well and watch for deadlocks though

 

Yes, but that's the same skillset as writing a scalable traditional web application.

CannotResolveSymbol said:

ManipUni said:

*snip*

The whole point of Azure is that it abstracts all that away from you.  You just write a plain old webapp; the Azure infrastructure takes care of all the hard parts:  data distribution/synchronization, automatic allocation of nodes, etc.

Wait, what? Where can I read more about this?

Is there an open source cloud OS being developed?  I can't see locking myself into both a hosting site and a proprietary OS. How do you pick up your application and run it elsewhere?

SteveRichter said:

Is there an open source cloud OS being developed?  I can't see locking myself into both a hosting site and a proprietary OS. How do you pick up your application and run it elsewhere?

Ubuntu has cloud alternatives.

SteveRichter said:

Is there an open source cloud OS being developed?  I can't see locking myself into both a hosting site and a proprietary OS. How do you pick up your application and run it elsewhere?

If you want to run your cloud applications on a non-proprietary OS, the only solution is to build it yourself. Even if a hosting provider builds a Linux-based cloud, there is no way they are going to let you mess with the code. It defeats the purpose of building out a common scaleable infrastructure.

kettch said:

SteveRichter said:

*snip*

If you want to run your cloud applications on a non-proprietary OS, the only solution is to build it yourself. Even if a hosting provider builds a Linux-based cloud, there is no way they are going to let you mess with the code. It defeats the purpose of building out a common scaleable infrastructure.

AFIAK Amazon lets you BYOOS (Bring Your Own OS).  Here is some more info on Ubuntu Server for Cloud Computing.

What's interesting about Ubuntu is exactly that you aren't locked into a specific hosting provider, in fact, you can easily set up your own "private cloud" or use someone like Amazon or Linode.

Bass said:

kettch said:

*snip*

AFIAK Amazon lets you BYOOS (Bring Your Own OS).  Here is some more info on Ubuntu Server for Cloud Computing.

 

What's interesting about Ubuntu is exactly that you aren't locked into a specific hosting provider, in fact, you can easily set up your own "private cloud" or use someone like Amazon or Linode.

I'd argue that the ability to have a custom virtual server/private cloud that is hosted somewhere is a lot different than the feature set that you get access too with something like Azure. Apples, meet oranges.

kettch said:

Bass said:

*snip*

I'd argue that the ability to have a custom virtual server/private cloud that is hosted somewhere is a lot different than the feature set that you get access too with something like Azure. Apples, meet oranges.

You also get a "feature set you get access too[sic]". You just are able to control what those "features" are.

This has to be the coolest offer from the worst looking website. Thanks for the link!

EDIT: It looks like the site has been updated.

-Josh

Bass said:

kettch said:

*snip*

You also get a "feature set you get access too[sic]". You just are able to control what those "features" are.

I guess I'm just thinking that it's a whole heck of a lot cheaper to pay somebody else to build the software that powers that infrastructure, but I'm also lazy like that. There are a lot of people who get their jollies saying that their stuff runs on an open source stack, when they will never actually take advantage of what open source means.

I'm still looking for some reason for using the cloud. In the meantime I'll just mention some things that could make it interesting.

1) Obv. the simple path (if it delivers) from LightSwitch to cloud sounds compelling if need arises.

2) A study (if not selling point) whether there is advantage in terms of lowering access latency of a web app/site (eg some chatty site/app, esp. when encrypted - I'm not sure why exactly but often they can be really slow to use), which is kinda hard to get with most traditional hosting providers who don't have capability to put a synchronized mirror of the whole site and all data behind every bottleneck connection around the world.

3) Earlier I asked about how DDOS proof is it and will you be forced to pay more if your site gets attacked in order to keep things online rather than MS considering it attack against the cloud platform and dealing with it without any fuss.

So three points which if the cloud platform could deliver without putting serious effort/money into, would be compelling. I don't know if they're important to whoever is already signing up to this, maybe MS has enough clients who don't care about these things for them not to matter? If they do matter I'd like to see studies, data, tests and so on and technical people talking on C9 how they implemented these. (Currently don't have enough interest in pouring through every released Cloud Cover etc video to see if these have been answered already, if someone knows these have been answered please link)