[erlang-questions] Lets build a stock exchange!
Thu Nov 15 13:55:22 CET 2007
>> How about a replacement for eBay auctions (or is Erlang what the e
>> mans now?-)? eBay has fault-tolerance, scale, distribution
>> A good solution would support 'live auctions', and keep the bids in
>> It'd have some of the challenges of a stock market, but you'd avoid
>> having to provide liquidity to make it useable (I realise your focus
>> is a book, but wouldn't you like millions of people to use the
> These problem domains lack one thing that makes a stock exchange such
> an interesting problem - the need for low latency! Building a large
> scale, distributed, fault-tolerant system is one thing - doing it
> while maintaining low latency is a whole different ball game. Showing
> that Erlang is well matched to these problems is a huge thing.
I agree that a good stock exchange solution implies low latency, but
I think a 'live auction' may need this too.
Further, I am not sure how low the latency will be; I can't ping
amazon, but when I ping google I get round-trip averages over 50 mSec
Revisiting the original post:
> I want you to start an Amazon EC2 instance and join the "Hardcore
> Erlang" cluster.
> I want to build the biggest Erlang cluster in the world and push
> Erlang to its limits.
My problem is I don't see why *I* would run a stock exchange
application, so I don't see why *I* would run Amazon EC2 instances.
On the other hand, I do see why I and others might want to run EC2
instances for auctions.
A plausible explanation for why I'd run the stock exchange
application would help me a lot.
Adjusting resources by starting and stopping EC2 instances also seems
to be a reasonable fit to auctions. Having a 'hub' which contains the
product catalogue would seem to be an enabler for the whole system,
and separate EC2 instances would be used for auctions (with some 'low-
rent' instances for low-activity auctions).
I am not 'wedded' to any particular application.
I *would* like to see Erlang running well on EC2, because that seems
to be great complementary technology.
I would also like to see something which would actually get run at
scale, and exploit the EC2 model, so that Erlang+EC2 gets proven.
More information about the erlang-questions