[Pharo-project] redundancy

Sven Van Caekenberghe sven at beta9.be
Sat May 14 11:55:26 CEST 2011


On 14 May 2011, at 09:54, Stéphane Ducasse wrote:

> Yes.
> So how do we do that?
> Dale can we use the ssgemstone you host?
> 
> Stef
> 
>> The problem is that the network operator is down...
>> 
>> see http://code.google.com/p/pharo/issues/detail?id=4218
>> 
>> We should really t hink hard about redundancy... 
>> 
> 

I think Amazon AWS S3, like Marcus suggested, is one way to go.

What if we build a bot (a Smalltalk image running somewhere) that periodically (say every 5 or 10 minutes) checks a specific number of important/critical Monticello compatible repositories and then syncs and if necessary copies versions to a properly structured (public readonly) S3 namespace.

S3 is just plain HTTP hosted data, so adding this to MC is already possible today, as a fallback like Lukas suggested.

Amazon S3’s standard storage is:
	• Backed with the Amazon S3 Service Level Agreement.
	• Designed to provide 99.999999999% durability and 99.99% availability of objects over a given year.
	• Designed to sustain the concurrent loss of data in two facilities.

It costs some money, but not much (it is pay as you go).

Sven

PS: Yes, I known the following question ;-)





More information about the Pharo-project mailing list