activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From magellings <mark.gelli...@qg.com>
Subject Re: [VOTE] Release Apache.NMS.ActiveMQ 1.2.0
Date Thu, 21 Jan 2010 18:41:49 GMT

It also should not be a failure if any one asynchronous connect fails.  It
should be if "all" asynchronous connect attempts fail.


magellings wrote:
> 
> Yes we use it.  This is an opt in feature.  The purpose of it is so on
> initial connect the client does not wait 20+ seconds for .NET to time out
> if it by chance attempts connecting to a slave broker first.  This is poor
> user experience.
> 
> 
> brandonagr wrote:
>> 
>> -1, has anyone else attempted to use the asynchronous connect for
>> failover? Nothing sets the Failure variable when there is a timeout so
>> there is a null reference exception if anything does timeout, at a
>> minimum there needs to be something setting the Failure variable:
>> 
>> Index: src/main/csharp/Transport/Failover/FailoverTransport.cs
>> ===================================================================
>> @@ -985,6 +985,7 @@
>>  
>>                              if(asyncConnect)
>>                              {
>> +								Failure = new NMSConnectionException("Async Timeout");
>>                                  Tracer.DebugFormat("Connect failed after
>> waiting for asynchronous callback.");
>>                              }
>> 
>> Also I'm not sure the approach is very good, if you have multiple broker
>> uris addressed it will fire off two threads and then wait for whichever
>> one connects first, this behavior makes specifying randomize=false
>> pointless as it will just connect to whatever responds first no matter
>> the order of the uris. It seems better that it would wait the async
>> timeout for one uri before attempting to connect to the next one.
>>  
>> 
>> 
>> 
>> Timothy Bish wrote:
>>> 
>>> Voting is open on Apache.NMS.ActiveMQ 1.2.0.
>>> 
>>> This release is based on the Apache.NMS API v1.2.0 and runs on .NET
>>> frameworks 2.0+ and Mono 2.0+
>>> 
>>> Changes in this version include
>>> 
>>> * Support for ConnectionMeteData in the Connection API.
>>> * Supports the new Individual Acknowledge Mode.
>>> * New IStreamMessage support
>>> * New IRedeliveryPolicy interface and PrefecthPolicy support.
>>> * Expanded IByteMessage interface to read/write primitive types.
>>> * Message's adhere to the JMS Read only and Write only rules.
>>> * Many new Unit Tests added to the Test Suite.
>>> * Support for Message Body Compression.
>>> * Connection Inactivity Monitor.
>>> * Optional Asynchronous Reconnects in the Failover Transport.
>>> * Many more bugfixes and improvements as well.
>>> 
>>> The binary and source bundles for the release candidate can be found
>>> here:
>>> <http://people.apache.org/~tabish/nms.activemq-1.2.0/>
>>> 
>>> The Wiki Page for this release is here:
>>> <http://activemq.apache.org/nms/activemq-net-120-release.html>
>>> 
>>> Please cast your votes:
>>> 
>>> [ ] +1 Release the source as Apache.NMS.ActiveMQ 1.2.0
>>> [ ] -1 Veto the release (provide specific comments)
>>> 
>>> Here's my +1
>>> 
>>> Regards,
>>> Tim
>>> 
>>> 
>>> 
>> 
>> 
> 
> 

-- 
View this message in context: http://old.nabble.com/-VOTE--Release-Apache.NMS.ActiveMQ-1.2.0-tp27211013p27262609.html
Sent from the ActiveMQ - Dev mailing list archive at Nabble.com.


Mime
View raw message