activemq-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Praveen Bysani <praveen.ii...@gmail.com>
Subject Re: Variable volume of messages across consumers
Date Tue, 23 Jul 2013 04:27:53 GMT
Hi Christian,

Thank you for the clarification. What are the factors in activemq that
could effect this, other than time to live that is set in the publisher ? I
have a time to live of 2 minutes currently, how can i ensure that most
messages are delivered before expiry ?

On 15 July 2013 20:04, Christian Posta <christian.posta@gmail.com> wrote:

> Yes, that's what it means.
>
> On Sunday, July 14, 2013, Praveen Bysani wrote:
>
> > Hi,
> >
> > I observed a lot of following messages in my log,
> >
> > 2013-07-15 03:48:22,399 | DEBUG | Message expired ActiveMQBytesMessage
> > {commandId = 1498, responseRequired = false, messageId =
> > d7dfc864-e46a-44b0-a0ec-b5d9e3a07127:1:1:253, originalDestination = null,
> > originalTransactionId = null, producerId = ID:global-desktop,
> destination =
> > topic://DM, transactionId = null, expiration = 1373831265580, timestamp =
> > 1373831145580, arrival = 0, brokerInTime = 1373831302399, brokerOutTime =
> > 1373831299130, correlationId = null, replyTo = null, persistent = false,
> > type = null, priority = 4, groupID = null, groupSequence = 0,
> > targetConsumerId = null, compressed = false, userID = null, content =
> > org.apache.activemq.util.ByteSequence@7f94dadb, marshalledProperties =
> > null, dataStructure = null, redeliveryCounter = 0, size = 0, properties =
> > null, readOnlyProperties = false, readOnlyBody = false, droppable =
> false}
> > ActiveMQBytesMessage{ bytesOut = null, dataOut = null, dataIn = null } |
> > org.apache.activemq.broker.region.RegionBroker | ActiveMQ Transport:
> > tcp:///<ip-address>:40887@6122
> >
> > Does that mean it failed to deliver the messages before the time to live
> ?
> >
> > On 12 July 2013 14:49, Praveen Bysani <praveen.iiith@gmail.com> wrote:
> >
> > > Hi,
> > > Thanks i added the log4j for all components under org.apache.activemq.
> > > Hopefully i can see something useful.
> > >
> > >
> > > On 12 July 2013 00:21, Christian Posta <christian.posta@gmail.com>
> > wrote:
> > >
> > >> Add logging configuration to turn on logging for the bridge.
> > >>
> > >> e.g.,
> > >>
> > >> to your log4j.properties, add
> > >>
> > >>
> > >>
> >
> log4j.logger.org.apache.activemq.network.DemandForwardingBridgeSupport=DEBUG
> > >>
> > >>
> > >> On Thu, Jul 11, 2013 at 3:38 AM, Praveen Bysani <
> > praveen.iiith@gmail.com
> > >> >wrote:
> > >>
> > >> > Hi,
> > >> >
> > >> > I found the following lines in the subscriber log
> > >> >
> > >> > | org.apache.activemq.transport.InactivityMonitor | DEBUG |
> 2013-07-11
> > >> > 07:34:44,384 | 30000 ms elapsed since last read check.
> > >> > | org.apache.activemq.transport.InactivityMonitor | DEBUG |
> 2013-07-11
> > >> > 07:34:44,384 |
> > >> > org.apache.activemq.transport.InactivityMonitor$2@61726a5c9999
> > >> > ms elapsed since last write check.
> > >> >
> > >> > I have disabled the inactivity period in the broker connection url
> by
> > >> > setting maxInactivityDuration to 0. Does it have anything to do with
> > >> > messages not being received ?
> > >> >
> > >> > On 11 July 2013 12:18, Praveen Bysani <praveen.iiith@gmail.com>
> > wrote:
> > >> >
> > >> > > Hi Christian,
> > >> > >
> > >> > > Thanks for your input. But i have little idea of what those terms
> > >> mean.
> > >> > > Could you elaborate please.
> > >> > >
> > >> > >
> > >> > > On 10 July 2013 21:06, Christian Posta <christian.posta@gmail.com
> >
> > >> > wrote:
> > >> > >
> > >> > >> Your best bet would be to set logging to debug on the network
> > bridge
> > >> > >> (DemandForwardingBridgeSupport) and possibly the RegionBroker
to
> > see
> > >> > what
> > >> > >> is happening.
> > >> > >>
> > >> > >> On Wednesday, July 10, 2013, Praveen Bysani wrote:
> > >> > >>
> > >> > >> > Hi,
> > >> > >> > I have the following setup in my project,
> > >> > >> >
> > >> > >> >    1. multiple producers (p1,p2,p3) in Hongkong region
sending
> > >> > variable
> > >> > >> >    size non-persistent messages to a topic t1 on the
broker
> (b1)
> > in
> > >> > >> >    Phillippines region
> > >> > >> >    2. p1,p2 and p3 send messages frequently (~1 second)
> > >> > >> >    3. each message has a time to live of 30 seconds
> > >> > >> >    4. a tcp network bridge from b1 to a new broker b2
in HK
> > region
> > >> > >> >    5. consumers c1 in hongkong and c2 in europe subscribed
to
> b2
> > >> > through
> > >> > >> >    ssl and a consumer c3 subscribed to t1 on b1 through
tcp
> > >> > >> >
> > >> > >> > I don't see the same amount of messages received across
c1, c2
> > and
> > >> c3.
> > >> > >> The
> > >> > >> > number of messages received is in the following order
c3 > c1 >
> > >> c2. I
> > >> > >> use
> > >> > >> > the default settings for prefetch values and *vmCursor
*for
> > >> > >> > *pendingSubscriberPolicy
> > >> > >> > *and *strictOrderDisptachpolicy*.
> > >> > >> >
> > >> > >> > While i understand there could be delays in passing
the message
> > to
> > >> > these
> > >> > >> > consumers as c3 could be slow consumer and c1 is fast,
> shouldn't
> > >> the
> > >> > >> > message eventually be received to all consumers of a
topic. Why
> > do
> > >> i
> > >> > see
> > >> > >> > different volumes of
>
>
>
> --
> *Christian Posta*
> http://www.christianposta.com/blog
> twitter: @christianposta
>



-- 
Regards,
Praveen Bysani
http://www.praveenbysani.com

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message