Database Reference
In-Depth Information
storm.messaging.netty.server_worker_threads: 1
storm.messaging.netty.client_worker_threads: 1
storm.messaging.netty.buffer_size: 5242880
storm.messaging.netty.max_retries: 100
storm.messaging.netty.max_wait_ms: 1000
storm.messaging.netty.min_wait_ms: 100
This change was actually much larger than simply adding Netty support.
Storm now technically supports the ability to add other sorts of transport
mechanisms.
Distributed Clusters
The default mode for Storm is to run as a distributed cluster of servers.
One machine serves as the host for the nimbus and, usually, the Storm
user interface (UI). Other machines run the supervisors and are added as
capacity is needed. This section describes the steps needed to start this
type of cluster, even if all the “machines” are actually running on the same
development box.
Starting the Servers
If this is a development machine, Storm contains a development ZooKeeper
install that can be run. You should start this before starting any of the
servers:
storm-0.9.0.1$ ./bin/storm dev-zookeeper
It is useful to keep this running in its own console as it logs out connections
and disconnections.
Now each of the servers can be started. On the machine acting as the
Nimbus, start the Nimbus server using the
bin/storm nimbus
command.
Additionally, start the web interface. It helps to manage the topologies:
storm-0.9.0.1$ ./bin/storm nimbus &
storm-0.9.0.1$ ./bin/storm ui &
Ofcourse,inaproductionenvironmenttheseshouldbeplacedintoastartup
script of some kind.
