qpid-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ted Ross <tr...@redhat.com>
Subject Re: [VOTE] Approve Qpid Dispatch 0.4-rc2 as the final 0.4 release
Date Thu, 09 Apr 2015 17:47:19 GMT

On 04/09/2015 01:39 PM, Gordon Sim wrote:
> On 04/09/2015 06:28 PM, Ted Ross wrote:
>> On 04/09/2015 01:20 PM, Gordon Sim wrote:
>>> On 04/06/2015 06:59 PM, Ted Ross wrote:
>>>> Here's the vote thread for Qpid Dispatch 0.4 (rc2) containing the fix
>>>> for the build problem Alan found in rc1.
>>>>      http://people.apache.org/~tross/qpid-dispatch-0.4rc2/
>>>> Please vote either to approve this candidate as 0.4 or report your
>>>> objections on this thread.
>>> Not an objection, just an observation... the helloworld_blocking.py
>>> example from proton doesn't work against dispatch. It appears that the
>>> router never sends back a disposition for the message sent by the
>>> client, and since the send() is blocking until settled, it can't
>>> progress to the receive().
>> This (or something similar) was discovered earlier and has been raised
>> in https://issues.apache.org/jira/browse/DISPATCH-111.  Ken, Alan, and I
>> discussed this and decided to defer that fix to the 0.5 release.
> I think that is a separate issue (from the description). Thinking about
> it some more the issue with the blocking helloworld is that there is a
> receiver link established for the message, but that link has no credit
> yet (because the credit is issued when calling receive()). In this case
> the router seems to not issue a disposition - it doesn't release because
> there is a link interested in the message; it can't accept because it
> hasn't received an accept from the receiver since it hasn't got credit
> to send to that receiver.
> Now that I've spelled it out for myself, this is certainly not a big
> issue at all as the example is not something you would ever really do.
> Annoying that it doesn't work, but I think perhaps better handled by
> rethinking the example.

I think you are correct.  The issue in DISPATCH-111 is related to 
link-routing which your test was probably not using.

Before you go and change the example, I'd like to look into this issue. 
  I think Dispatch should be as transparent as possible to things like 
this (even if they're not _good_ examples).

> +1 on 0.4 from me.


> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@qpid.apache.org
> For additional commands, e-mail: users-help@qpid.apache.org

To unsubscribe, e-mail: users-unsubscribe@qpid.apache.org
For additional commands, e-mail: users-help@qpid.apache.org

View raw message