2.1.1 System configuration
The following figure shows a system configuration.
|
- Organization of this subsection
(1) Execution environment
This is the environment you need in order to use the application programs that access the distributed KVS, as well as the distributed KVS itself.
You must use a reliable network. To improve reliability, we recommend that you use dual networks between EADS servers.
(a) EADS server
EADS server refers to a server process that manages data consisting of keys and values.
The following is the program product used to configure EADS servers:
-
Hitachi Elastic Application Data Store
(b) Cluster
Normally, a system consists of multiple EADS servers. A group of EADS servers is called a cluster. EADS clients recognize the cluster as a single unit of storage.
A cluster consists of a group of EADS servers that have the same multicast address and port number within the same segment.
- Important note
-
The number of EADS servers that make up the cluster must be at least the data multiplicity (the number of redundant copies of data plus the original)
2 - 1.
For details about the number of redundant copies of data, see 2.8 Creating redundant copies of data.
(c) EADS client
EADS clients refer to user programs that use client libraries provided by EADS to connect to EADS servers.
There are two types of EADS clients, corresponding to the programming languages used to create application programs (Java or C):
-
EADS client (Java)
Hitachi Elastic Application Data Store Client for Java
-
EADS client (C)
Hitachi Elastic Application Data Store Client for C
- Important note
-
EADS does not support a system configuration that involves conversion of IP addresses or port numbers for communication between EADS clients and EADS servers.