vrs-development
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Vrs-development] pnetlib and the VRS, SEE and other server concepts


From: Chris Smith
Subject: [Vrs-development] pnetlib and the VRS, SEE and other server concepts
Date: Tue, 19 Mar 2002 17:35:07 +0000

I've picked up on a few Socket layer related threads recently, which leads me 
to ask exactly how a C# webservice deployed over pnetlib is accessed? (in a 
network sense).

Without going into the whole VRS thing because this applies to all network 
server concepts..... I've been visualising some sort of server that receives 
a webService request (say SOAP), decodes the request, discovers the 
webService that needs to be invoked, brings up the VM with the webservice in 
tow and passes it the request data.

Now of course this webservice may call another webservice, which is 
definitely where I see the pnetlib support coming in.

So I've always imagined the service request capture and decode being external 
to the actual web service itself.  Is this correct?

Secondly, with something like the VRS, which will be tightly bound to the VM 
for sandbox reasons, will it be possible to surp the low-level architecture 
specific layer of the pnetlib socket class implementation (as well as the 
file I/O layer) into VRS calls so that the VRS can determine whether a call 
across the network is required or can be satisfied locally?

Actually that's not quite right is it? because what we'd really be looking 
for is a 'call webService' class/method which we could surp - hitting the low 
level socket layer isn't a help so I take back what I just said! :o)
There isn't such a class is there??? :o(

However, surp'ing the io layer for files etc IS important for the VRS to 
route file access into the 'virtual file system' instead of the real OS 
filesystem.


I don't think pnet and deployment servers are going to meet on friendly terms 
when we start to draw everything together........
Has anyone thought about this?

Are webservices really going to do their own request decoding/reply packing???

I'd like to see a system where the delivery and content encapsulation of 
requests is transparent and handled by a netserver component, and webservices 
just get the data payload (XML of course! :o)
Delivery of requests can then be SOAP etc over HTTP/SMTP/FTP whatever.

Any thoughts?  I need to get a handle on this!

Cheers
Chris

-- 
Chris Smith
  Technical Architect - netFluid Technology Limited.
  "Internet Technologies, Distributed Systems and Tuxedo Consultancy"
  E: address@hidden  W: http://www.nfluid.co.uk



reply via email to

[Prev in Thread] Current Thread [Next in Thread]