On Fri, 14 Dec 2007 09:43:12 +0100 jean-frederic clere <[EMAIL PROTECTED]> wrote:
> > But I still don't understand why proxy which has an option to get > > proper data from another (probably) working server stops trying to > > failover. I think from proxy point of view it doesn't matter why > > server is unavailable. Or am I missing something? > > The ajp proxy may have sent the request and get a partial response if > you fail over after HTTP_INTERNAL_SERVER_ERROR you may duplicate a > transaction. That argues that we might make want to identify idempotent requests and treat them differently. That's going to get a bit hairy: we may be using HTTP status codes in inconsistent ways here: - as errors propagated from a backend. - as errors generated in the proxy "XYZ is unrecoverable". - as HTTP representations of errors from another protocol. I think the mod_proxy scheme handlers would need to decide whether a backend request can be retried, and return different statuses for when it is/isn't. That's what it does with the INTERNAL_SERVER_ERROR vs SERVICE_UNAVAILABLE distinction, albeit that's not really adequate if it's doubling up as what we return to the client. If the OP has in mind an improved scheme, perhaps he should think it through to the point where he can hack up a proposal? -- Nick Kew Application Development with Apache - the Apache Modules Book http://www.apachetutor.org/
