Hi Haritha- The broker network works to optimize performance and also prevent loops. After a number of transfers between brokers, the broker will stop sending messages to prevent looping.
This sounds like a common problem if you are using Spring JMS and/or Apache Camel JMS. The problem its that he connections are opening and closing very quickly and this causes a lot of noise to the broker network. The solution is to use a PooledConnectionFactory to keep connections open and to try to use CACHE_CONSUMER when possible to prevent consumer thrashing (and subsequent network looping). Thanks, Matt Pavlovih > On Feb 2, 2023, at 11:03 AM, Justin Bertram <jbert...@apache.org> wrote: > > What version of the broker are you using? What kinds of clients are you > using? How are your brokers configured? What is the distribution of > consumers across the queues in the cluster? > > > Justin > > On Thu, Feb 2, 2023 at 10:57 AM Haritha Desikan <hdesi...@spscommerce.com> > wrote: > >> Hi, >> >> Our company recently decided to make a shift to Active MQ. Our >> infrastructure currently uses a mesh of three active/standby broker pairs >> and we have been facing an issue with the brokers where redelivered >> messages sitting on a queue in one particular broker are not being >> transported to the broker with active consumers on it. >> >> Please provide some guidance/insight as to what can be done in such >> scenarios. >> >> Thank you, >> Haritha >> Cloud Systems Engineer >> SPS Commerce >> >>
