[FoRK] overhead of RESTful stuff

Stephen Williams sdw at lig.net
Thu Mar 8 16:09:11 PST 2012

On 3/8/12 3:53 PM, J. Andrew Rogers wrote:
> On Mar 8, 2012, at 4:25 AM, Eugen Leitl wrote:
> ...
>> In terms of networking, is GBit Ethernet going to become a
>> bottleneck for building applications from REST interfaces?
> Not likely. You will run into other bottleneck's first if you are using REST interfaces. If efficiency, scaling, and performance is your primary concern, REST is not what you are looking for most days. The exception might be if you are moving large amounts of data bidirectionally for each operation, in which case you will be bandwidth limited.
>> Are any of you using REST on InfiniBand or is the idea insane
>> in general, given the (probably awful) overhead?
> Insane. REST is terrible for high-performance network communication. Using Infiniband for REST would be like putting a Ferrari engine in a Trabant.

Although the HyBi guys are getting close to a solution, it is sad that we've been stuck with synchronous RPC-ish HTTP.  Async 
pipelined message queues with true bidirectional message processing with few open connections would be far better.  And that's what 

You can do REST principles on that, modified in various ways for distributed messaging topology.  EXI/Protocol Bufers with MPI 
capabilities and W3C/OAuth addressing and security...

But few people care.  Most scalable things have enough money to buy thousands of servers and run junk code.

> --
> J. Andrew Rogers
> Twitter: @jandrewrogers

More information about the FoRK mailing list