Does the broker put the full stack trace of the NullPointerException in its
logs?

I'll try to fix this (as well as the out of order detach) if I can locate
it.

Cheers,
Rob

On 3 December 2015 at 10:03, Gordon Sim <[email protected]> wrote:

> On 12/02/2015 09:45 PM, Rob Godfrey wrote:
>
>> On 2 December 2015 at 21:45, Gordon Sim <[email protected]> wrote:
>>
>>>  From the proton python examples, I was unable to connect as that client
>>> populates the hostname with a host and port combination, which was
>>> causing
>>> the broker to close the connection:
>>>
>> [...]
>
>> In its default configuration the Java Broker will accept any valid
>> hostname
>> or IP address as the "hostname", or the empty string, or "default".  It
>> will not match anything containing the port number, nor will it match to
>> 0.0.0.0.  To get the Python client to work you could add the following
>> entry:
>>
>> {
>>        "name" : "pattern",
>>        "virtualHostNode" : "default",
>>        "type" : "patternMatchingAlias",
>>        "pattern" : ".*"
>> }
>>
>> into the list of virtualhostaliases within the AMQP port entry within the
>> config.json file.  This will match any hostname in the open frame to the
>> virtual host named "default".
>>
>
> Yes indeed, that works perfectly, thanks! With that in place most of the
> python examples work as expected. The one exception is those involving the
> server, which creates a sending link with no address in the target (the so
> called 'anonymous-relay' behaviour I believe?). The java broker responds
> with:
>
> [0x8c9450]:0 <- @close(24) [error=@error(29)
>> [condition=:"amqp:connection:forced",
>> description="java.lang.NullPointerException"]]
>>
>
> That of course isn't an issue per-se as the behaviour in question isn't in
> any way official, its just a useful convention. I mention it here only in
> case it is expected to work and there is some minor interop issue we can
> fix (and also for future reference if anyone else hits it).
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>

Reply via email to