I set up a network of brokers with a topology as shown below.
- Manufacturer (P1) connected to Broker (B1) and Producer (P2) connected to Broker (B2)
- Broker (B1) and Broker (B2) are connected as a network of brokerage companies and are balanced by balance.
- Consumer (C1) connected to the Broker (B1) and Consumer (C2) connected to the Broker (B2)
Clients are configured to use fault tolerance as:
- Consumer-1 = switching to another resource: tcp: // localhost: 61616, tcp: // localhost: 61615? randomize = false
- Consumer-2 = failover: tcp: // localhost: 61615, tcp: // localhost: 61616? randomize = false
As soon as channel-2 goes down, P2 and C2 will move to channel-1, which is the desired failure behavior.
I want to understand the behavior when the Chaneel-2 returns? I noticed that only Channel-1 continues to service all connections even after channel-2 has recovered and, thus, has lost load balancing between the channels.
I want to know if it is possible, as soon as Channel-2 returns, load balancing will start automatically between the channels, and the corresponding Producer-2, Consumers-2 will shift to Channel-2 and, thus, will ensure full load balancing and a complete transition to another resource ?
I came across the article “Combining Failover with Load Balancing” at http://fusesource.com/docs/broker/5.4/clustering/index.html is recommended for combining failover and load balancing?
Hello,
-Amber
source share