Another piece of information: the exact exception stack is as follow: [Dummy.Framework.Notification.ActiveMQ.Tools.TopicConnection.OnExceptionListener] [5] System.IO.EndOfStreamException: Unable to read beyond the end of the stream.
at System.IO.__Error.EndOfFile() at System.IO.BinaryReader.FillBuffer(Int32 numBytes) at System.IO.BinaryReader.ReadInt32() at Apache.NMS.Util.EndianBinaryReader.ReadInt32() at Apache.NMS.ActiveMQ.OpenWire.OpenWireFormat.Unmarshal(BinaryReader dis) at Apache.NMS.ActiveMQ.Transport.Tcp.TcpTransport.ReadLoop() 0 2010/1/26 Sebastian Rodriguez <srodrig...@gmail.com> > From the log messages I can see: > > 1/25/2010 9:49:09 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Warn] > [6] ActiveMQ.NMS.Warning: Transport failed to > tcp://localhost:61616/, attempting to automatically reconnect due to: Unable > to read beyond the end of the stream. 0 > > 1/25/2010 9:49:09 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [6] ActiveMQ.NMS.Debug: Transport has been > Interrupted. 0 > > 1/25/2010 9:49:09 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Attempting connect to: > tcp://localhost:61616/ 0 > > 1/25/2010 9:49:09 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Attempting connect to: > tcp://localhost:61616/ 0 > > 1/25/2010 9:49:09 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Opening socket to: localhost on port: > 61616 0 > > 1/25/2010 9:49:10 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Connect fail to: > tcp://localhost:61616/, reason: Error connecting to localhost:61616. > 0 > > 1/25/2010 9:49:10 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Waiting 10ms before attempting > connection. 0 > > 1/25/2010 9:49:10 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Attempting connect to: > tcp://localhost:61616/ 0 > > 1/25/2010 9:49:10 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Attempting connect to: > tcp://localhost:61616/ 0 > > 1/25/2010 9:49:10 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Opening socket to: localhost on port: > 61616 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Connect fail to: > tcp://localhost:61616/, reason: Error connecting to localhost:61616. > 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Waiting 20ms before attempting > connection. 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Attempting connect to: > tcp://localhost:61616/ 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Attempting connect to: > tcp://localhost:61616/ 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Opening socket to: localhost on port: > 61616 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Connected to localhost:61616 using > InterNetwork protocol. 0 > > 1/25/2010 9:49:12 AM Debug: [ > Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Info] > [7] ActiveMQ.NMS.Info: Restoring previous transport > connection. 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [6] ActiveMQ.NMS.Debug: Parsing type: 1 with: > Apache.NMS.ActiveMQ.OpenWire.V1.WireFormatInfoMarshaller 0 > > 1/25/2010 9:49:12 AM Debug: [ > Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Info] > [7] ActiveMQ.NMS.Info: Sending queued commands... 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Transport has resumed normal > operation. 0 > > 1/25/2010 9:49:12 AM Error: > [Dummy.Framework.Notification.ActiveMQ.Tools.TopicConnection.OnConnectionResumedListener] > [7] ACTIVEMQ: Connection with Broker Resumed 0 > > 1/25/2010 9:49:12 AM Debug: > [Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Debug] > [7] ActiveMQ.NMS.Debug: Connection established > 0 > > 1/25/2010 9:49:12 AM Debug: [ > Dummy.Framework.Notification.ActiveMQ.Tools.ActiveMQTracer.Info] > [7] ActiveMQ.NMS.Info: Successfully reconnected to: > tcp://localhost:61616/ 0 > Is it the normal behavior that we do not receive a ConnectionInterrupted > event? > > What can be the causes to receive the exception Unable to read beyond the > end of the stream? I've seen a jira speaking about it ( > http://issues.apache.org/activemq/browse/AMQNET-196) but i do not receive > this exception in a regular basis. It seems that this exception occurs more > often in slow or old machines. Can it be related? > > Thank you very much for your help on the subject, > > Sebastien > 2010/1/25 Sebastian Rodriguez <srodrig...@gmail.com> > > I've enabled the logging and left the client running during the night. >> Hopefully I'll be able to come back to you tomorrow with more details. >> The "funny" thing, is that there is no connection lost, we "only" looses >> the subscriptions. >> I'll try to make a sample tomorrow see if i can reproduce the error >> >> 2010/1/25 Timothy Bish <tabish...@gmail.com> >> >>> On Mon, 2010-01-25 at 12:18 +0800, Sebastian Rodriguez wrote: >>> >>> > Dear all, >>> > I'm encountering a very strange problem. >>> > I have a client that uses an embedded broker to connect to a network of >>> > brokers. >>> > The connection between my client and the embedded broker is done >>> through the >>> > Apache.NMS.ActiveMQ client. >>> > I'm connecting to the embedded broker through a failover uri as I want >>> to >>> > recover from possible exception raised. >>> > The version of the NMS client is 1.2.0.1794 and we are using the latest >>> > version from 5.3.1 branch. This was also happening with the stable >>> release >>> > 5.3.0. >>> > >>> > Everything works perfectly until I receive the event >>> > ConnectionResumedListener. >>> > The weird part is that: >>> > >>> > - why do I receive a connection resume event without a previous >>> > connection interrupted event? >>> > - the connection is recovered properly, but not my subscription. >>> > Basically my client doesnt receive anything anymore from the >>> embedded broker >>> > although the broker still receives the messages incoming from the >>> network of >>> > brokers. >>> > >>> > Do you have any ideas why I would receive a connection resumed alone >>> > (without a connection interrupted?), and why the failover transport >>> > correctly recovers the connection but not subscriptions? >>> > Thank you very much for your advice on the matter, >>> > >>> >>> Can you enable logging and capture some logs around the time the event >>> is fired? I have not seen any spurious disconnects as yet from NMS. If >>> you have a sample app that can demonstrate the problem that would be >>> helpful. >>> >>> Regards >>> Tim. >>> >>> >>> >>> >>> >>> >>> -- >>> Tim Bish >>> http://fusesource.com >>> http://timbish.blogspot.com/ >>> >>> >>> >>> >> >> >> -- >> Sebastien Rodriguez >> > > > > -- > Sebastien Rodriguez > -- Sebastien Rodriguez
