OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

wsrf message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Cross Posting - [ogsa-wg] How to make effective use of WS-BaseFault in GFF specs


Folks,

More from Steve on Base Faults.

Begin forwarded message:

> From: Steve Loughran <steve_loughran@hpl.hp.com>
> Date: 17 January 2005 17:13:30 GMT
> To: David Snelling <David.Snelling@uk.fujitsu.com>
> Subject: Re: [ogsa-wg] How to make effective use of WS-BaseFault in 
> GFF specs
>
> David Snelling wrote:
> > Steve,
> >
> > There has not been much guidance of the use of BF done in public. You
> > might look at the Globus alpha release.
> >
> > Also a couple of these look like good issues for the WSRF-BF issue 
> list.
> > Can I cross post it there?
>
> Yes, they can even have a look at the relevant bit of the latest draft 
> of the deployment API, which I have attached.
>
> Ideally it would be nice for there to be a well known 'Wrapped SOAP 
> Fault' class. Same even for a WrappedHttpFault, where things like a 
> 304 redirect with the redirection header would be included, but that 
> needs a SOAP stack that goes down to that level. We added the 
> HttpFaultcode in Axis to track down low level problems (esp with 
> proxies and wrong URLs), but there are a lot more diagnostics that 
> could be captured at that point.
>
> The use case for some of the fields is trying to track down problems 
> on your server cluster with nothing but a customer at the end of a 
> telephone to provide diagnostics. The more  diagnostics, the better, 
> both client side and server side. As a personal example, one problem 
> we tracked down on client complaints of "your service is returning bad 
> XML" was really "your client is pointing at the wrong URL"; the client 
> was trying to parse the textual 404 error page as if it was an XML 
> message, and throwing SAX errors...
>
> That was more amusing than when DNS failed on one node causing its 
> requests to fail early, with the load balancing router then directing 
> more traffic to that node because its throughput was higher.
>
> see: http://www.iseran.com/Steve/papers/when_web_services_go_bad.pdf
>
> -steve
>
>
>
>

faults.pdf

>
-- 

Take care:

     Dr. David Snelling < David . Snelling . UK . Fujitsu . com >
     Fujitsu Laboratories of Europe
     Hayes Park Central
     Hayes End Road
     Hayes, Middlesex  UB4 8FE

     +44-208-606-4649 (Office)
     +44-208-606-4539 (Fax)
     +44-7768-807526  (Mobile)



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]