Rodeo: 2 questions

Pierre Sahores psahores at free.fr
Thu May 20 13:25:32 EDT 2010


Years ago, i did test an XServe G4 running a Sybase ASE 12.5 and went able to get 1500 web served requests / secs without any server's stress at all. And as anyone should know, OS X Server is lots less responsive than Linux or BSD.

2 cents,

Best, Pierre

Le 20 mai 2010 à 18:49, Andre Garzia a écrit :

> David,
> 
> I don't think you'll reach problems of scalability that easily. Twitter and
> Facebook have scalability issues, you'll probably be fine for months before
> reaching scaling issues even if you're really successful.
> 
> Don't think a single server with a single database is no good for your
> needs. Can you tell me what is the biggest demand you think of for your
> product because I tend to believe that you'll do just fine with a simple
> setup.
> 
> You need massive access and stuff to approach the limits of mySQL or
> PostgreSQL. Apache is very robust as well and I don't think you'll reach its
> limit.
> 
> You're probably safe on On-Rev or Rodeo or whatever is invented soon.
> 
> Andre
> 
> On Thu, May 20, 2010 at 1:25 PM, David Bovill <david at vaudevillecourt.tv>wrote:
> 
>> On 20 May 2010 16:55, Jerry Daniels <jerry.daniels at me.com> wrote:
>> 
>>> The cheapest, most scalable and fastest performing are all the same
>>> solution:
>>> 
>>> 1. Client: thin
>>> 2. Web server: thin, but round-robin'd the IP addresses to 1 of the 13
>> app
>>> servers
>>> 3. Web app server: hefty, almost fat
>>> 4. Data: thin and agnostic (NO stored procedures)
>>> 
>> 
>> Hi Jerry this is not the sort of scalability that is needed for some
>> interesting classes of apps. First it is very expensive in terms of set up,
>> and then admin. By very expensive I mean more than $1,000.
>> 
>> It is the transition between - "give the idea a go" and "wow it's taken
>> off"
>> that I'm interested in addressing. If you can get the costs down on that
>> you
>> can do some interesting things. At the progression from basic hosting to
>> the
>> set up you describe is a big expensive jump. Also it does not scale
>> massively for bursts on unpredictable demand. One application I've been
>> asked to get my head around may have up to 1 million concurrent users or it
>> may flop - a pay as you go service like Amazon or Google App engine helps
>> you cope with that.
>> 
>> In the world of webApps, I think we can also consider other scenarios:
>> 
>> 
>>  1. AJAX embeds / Flash / revLet plugins for blogs, webApps on mobiles
>>  2. Client side processing and web service based data => no need for 2)
>>  3. Cloud based DB such as Google AppEngine or Amazon SimpleDB
>>  (effectively combines 3 and 4)
>> 
>> People buy the apps, come to a separate web site where they can create
>> customised embeds for their blogs or social networks. They can buy or
>> subscribe and this covers the cost of the Cloud DB as it scales
>> _______________________________________________
>> use-revolution mailing list
>> use-revolution at lists.runrev.com
>> Please visit this url to subscribe, unsubscribe and manage your
>> subscription preferences:
>> http://lists.runrev.com/mailman/listinfo/use-revolution
>> 
> 
> 
> 
> -- 
> http://www.andregarzia.com All We Do Is Code.
> _______________________________________________
> use-revolution mailing list
> use-revolution at lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-revolution
> 

--
Pierre Sahores
mobile : (33) 6 03 95 77 70

www.woooooooords.com
www.sahores-conseil.com









More information about the use-livecode mailing list