Skip Menu |

This queue is for tickets about the AnyEvent-RabbitMQ CPAN distribution.

Report information
The Basics
Id: 97716
Status: resolved
Priority: 0/
Queue: AnyEvent-RabbitMQ

People
Owner: DLAMBLEY [...] cpan.org
Requestors: carl [...] cloudamqp.com
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: (no value)



Subject: Respect channelMax
Date: Sun, 3 Aug 2014 12:31:00 +0200
To: bug-AnyEvent-RabbitMQ [...] rt.cpan.org
From: Carl Hörberg <carl [...] cloudamqp.com>
Starting with RabbitMQ 3.3.5 the server will disconnect connections that don't respect the channelMax connection tune property. More details: https://groups.google.com/forum/#!topic/rabbitmq-users/Q6TBGjqEBpQ It looks like this lib correctly supports channelMax, but the uint16 is overflowing here: https://metacpan.org/source/BOBTFISH/AnyEvent-RabbitMQ-1.17/lib/AnyEvent/RabbitMQ.pm#L47 subtract one and you should be good.
The AMQP 0.9.1 spec doesn't make clear whether or not "short" is signed or unsigned. Would it be safer to default the max to 32767? https://www.rabbitmq.com/specification.html
Subject: Re: [rt.cpan.org #97716] Respect channelMax
Date: Mon, 23 Mar 2015 15:28:09 +0800
To: bug-AnyEvent-RabbitMQ [...] rt.cpan.org
From: Carl Hörberg <carl [...] cloudamqp.com>
In RabbitMQ it's unsigned, max allowed value is 65535.. On Sunday 22 March 2015 at 01:54, Dave Lambley via RT wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=97716 > > > The AMQP 0.9.1 spec doesn't make clear whether or not "short" is signed or unsigned. Would it be safer to default the max to 32767? > > https://www.rabbitmq.com/specification.html
On Mon Mar 23 03:28:25 2015, carl@cloudamqp.com wrote: Show quoted text
> In RabbitMQ it's unsigned, max allowed value is 65535..
Apologies for returning to this after so long! I have subtracted 1 as suggested. https://github.com/bobtfish/AnyEvent-RabbitMQ/pull/28 Best regards, Dave