You can certainly embed Artemis in your application. Then you can send messages to the local broker and then configure a bridge to forward the messages to the remote broker. The bridge will deal with the network outages while your clients will be free to send messages to the local broker without issue. Assuming the messages are durable then you will be able to power off your client and when you power it back on any messages that hadn't already been forwarded to the remote broker will still be there and will be forwarded when the network allows.
To be clear, Artemis is under consistent development but most of that work is to achieve more parity with ActiveMQ 5.x specific features (e.g. OpenWire support), JAAS login modules, etc. The core broker hasn't change much in a long while. It's stable and hardened and used in lots of production situations. It should suit your use-case just fine. Justin ----- Original Message ----- From: "Alessandro Manzoni" <manzoni.alessand...@gmail.com> To: users@activemq.apache.org Sent: Saturday, April 16, 2016 4:40:58 AM Subject: Fwd: How to Hi, I'm new to artemis activemq, so be patient. I'm facing a project that requires to exchange messages. OK, this is a MOM job. But one of project requirements is to enable lots of clients to send messages over a network that intermitently goes down for some while, and when the network resumes all generated messages should deliver to the proper destination, even if meanwhile the client has been powered off and on again. I know tha I could do this IBM MQ, but due to budget restrictions I would like to use ActiveMQ Artemis. I hoped to be able to embed a broker into the client software capable to communicate with a central broker as a connection is available, but I read on "Apache ActiveMQ Artemis User Manual" manual somhere, that clients won't be connecting to an embedded broker. Isn't it? Do I succeed achieving my goals using ActiveMQ Artemis? How to? Thanks.
