jsr339-experts@jax-rs-spec.java.net

[jsr339-experts] Re: MessageTooLargeException: status 413

From: Sergey Beryozkin <sberyozkin_at_talend.com>
Date: Wed, 5 Dec 2012 15:24:03 +0000

On 05/12/12 15:13, Bill Burke wrote:
>
>
> On 12/4/2012 12:00 PM, Sergey Beryozkin wrote:
>> I thought I would not worry about the related enhancement request being
>> rejected, but I'm changing my mind :-) - I'm still easy on it, but
>> sending this message to the list in case someone may also find an idea
>> of introducing an exception class representing HTTP 413 useful.
>>
>> While I agree with Marek that we have to be careful about not
>> introducing an exception class for every possible HTTP status starting
>> from 300, I think MessageTooLargeException is the last missing exception
>> class which can actually be practically used.
>>
>
> I actually disagree with the notion we can't have an exception class for
> every HTTP status code where it makes sense for. What is the harm?
>
Personally I do not see any particular harm; but - I can think of
why/when any of the already added exception classes can be used. If no
other status covered specifically covered yet then I guess we haven't
had any particular demand for such specific exceptions yet

>> Specifically, CXF reports 413 when the multipart body is too large, as
>> well in some other cases. I can imagine users of jersey and Resteasy
>> also finding it useful...
>>
>
> We use mime4j for multipart and I believe it will save large bodies to a
> temporary file.

Yes - but I'm referring to the runtime effectively enforcing the denial
of service attacks or preventing the overload where the user has
explicitly requested so, limit the payload to a specific max size only...


> But, IMO, we should have an exception for every error
> code anyways (and even some redirects).
>
>

Cheers, Sergey