[EMAIL PROTECTED] wrote: > Shawn, > >> webrev: >> http://cr.opensolaris.org/~swalker/pkg-3014/ > > I'm not familiar with most of this code, so I'm afraid I don't have a > lot of useful feedback to provide. > > I did have an error handling nit. > > face.py: > > - respond() and __handle_error() both have the potential to return a > 500 error. In respond, an IOError that isn't EISDIR will generate a > 500 error. It seems like we should be more careful, especially in > this case. There are a bunch of environment errors that probably > aren't 500s, permissions are a good example.
I could see that. > In both cases, however, returning a 500 will cause Apache's > mod_balancer to take us out of the pool. Can we return another > error here, or at least reduce the instances where we might get a > 500 error? I'm not sure what to do here. 500 is the correct response status here, but I realise that Apache is sort of important :-) However, RFC2616 [1] implies that you can use a 404 code instead any time you don't want to reveal the real reason a request failed [2]. I'd like Dan or Brad's input on this as they have dealt the most with this issue. Thanks for the feedback, -- Shawn Walker [1] http://www.w3.org/Protocols/rfc2616/ [2] http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html#sec10.4.5 _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
