ACTIVEMQ Official Document Analysis __ACTIVEMQ

ACTIVEMQ Real-time features:

Multicast: Activeblaze delivers fast, reliable point-to-point messaging of infrastructure to meet the requirements of high-performance and collaborative applications. Using a basic building block of a channel, there is a functional hierarchy for different problem areas, one for each superset. Blazechannel supports broadcast themes-using reliable multicasting or point-cast;

Group membership: Activeblaze Support Group members use groupchannel--This can be created from groupchannelfactory. You must explicitly say which group you want to join/leave, you can join multiple groups. Each groupchannel has both a global unique ID and a name.

Lifecycle: Channel is one of five: constructs-the channel is not initialized or is closed

Activeblaze message type: The basis for message and event resolution Activeblaze Blazemessage. Blazemessage is a mapped key-value pair, where the key is in the string and the value of the original object.

Broker Camel Component

Embedded in the Apache ACTIVEMQ Camel Agent provides great flexibility to expand the message broker camel's integration force. The Apache Camel route also benefits from remote connectivity to ACTIVEMQ serialization and network costs--if you use the ACTIVEMQ component cluster cluster is a big theme that usually means different things to different people. We will try to list all aspects of the cluster and the messages they have with the Activemq Queue consumer CLUSTERS:ACTIVEMQ support reliable high performance load balancing on the consumer queue. In enterprise integration, this situation is known as the competitive consumer model. The following figure illustrates this concept.

This solution sends the receiving message to the producer, who can allocate all the registered users between them. This has many advantages: the load is distributed in a very dynamic fashion. Additional consumers can supply and connect to queues during periods of high load without modifying any configured queues, for new consumers will show competition for another consumer. Better availability than the system uses a load balancer. Load balancers often rely on the monitorization system to find out which real servers are unavailable. With competing consumers, consumers will not compete for failure messages, so they will not deliver messages even without monitorization. High reliability, if consumption fails, any unrecognized queue of messages is sent to other consumers.

The disadvantage is that this model may not be ideal for system order processing that is required. The consumer competition model should be used in conjunction with other ACTIVEMQ features to mitigate this problem, while maintaining the benefits of using an exclusive consumer and ACTIVEMQ FAQ from the messaging group.
Broker clusters

The most common mental model clustering in a JMS context is a collection of JMS proxies and a JMS client will connect to one of them; if the JMS agent drops, it will automatically reconnect to another agent. We implement this use failover://JMS Client protocol. For more information, see How to configure a failover protocol for a Failover transport reference page. NOTE: Reliable://Agreement ACTIVEMQ 3. X has been changed to failover://protocol failover://
If we just run multiple agents on the network, tell the customer to use the discovery static or dynamic discovery, and then the customer can easily failover from one agent to another. However, independent brokers are unaware of consumers in other brokers; if there is no consumer agent, information can accumulate without being consumed. We have a salient feature that requires solving this problem at the client, but the current solution to this problem is to create a broker between the store and forward messages between network brokers.
Discovery of brokers

We support automatic discovery of discovery brokers using static or dynamic discovery, so customers can automatically detect and connect to the broker's logical group of brokers and brokerage firms to discover and connect to other agents to form large networks.
Networks of brokers

If you are using a client/server or hub/spoke style topology and you have a lot of clients and brokers, there is an opportunity for a proxy manufacturer, but no consumer, this message is stacked without being processed. In order to avoid this situation, ACTIVEMQ supports network brokers to provide storage and forwarding with consumers and producers to send messages from brokers to brokers that allow us to support distributed queues and topics through the network. This allows the client to connect to any agent, failing over to another agent if there is a failure--providing a cluster of brokers from the customer's perspective. The broker also allows us to expand the network of large numbers of clients we can run we need as many brokers as possible. You can take this as a cluster of clients to contact a cluster broker Auto-failover and discover, do a simple and easy-to-use message structure.

Master Slave

Run a large number of independent broker issues or brokers in a network message belonging to a single physical agent at any time. If the agent drops, you must wait for it to reboot before the message can be delivered. (If you are using non-persistent messages and the broker drops usually you lose your message). The idea behind the master is that the message is copied to a slave agent, so even if you have a catastrophic master machine hardware failure, file system or data center, get a direct failover to the slave without losing any information.
Another option for replicated message Stores is to have the means to replicate messages stored, so to some extent shared disk files. For example, using a SAN or shared network drive can share a file's proxy, so that if another agent fails to take over directly. So by supporting replication message storage you can reduce the risk of information loss provide a full HA backup or datacenter solution to survive Dr. Failure