Clustering Plus

Photos are courtesy of Liang Hong

Products

TPS Clustering PlusTM

TPS Clustering Plus is an innovative clustering solution that delivers superior performance, and ease of use at an affordable price. It keeps mission-critical information available and accessible on multiple network locations, and also protects users from being disconnected from data if one or two network servers become unavailable.

Key Benefits

Improves IT investment by maximizing utilization of existing hardware and software assets
Enhances business continuity by preventing impact from unforeseen disaster conditions
Deploys easily in heterogeneous server and storage environments
Delivers better service by reducing network traffic
Simplifies network topology and system management

Product Features

TPS Clustering Plus is built on innovative Mirror File System technology whose unique file-level operational mechanism replicates files and directories over the network in real time. It enables:

Active-Passive Setup Support: In contrast to conventional active-standby clustering setups, TPS Clustering Plus supports a writable active server and readable passive server. This eliminates the need for the standby or unused resources required by existing active-standby clustering models. The passive server can be any UNIX or Linux server that supports the NFSv2 or NFSv3 protocol.
Active-Active Setup Support: TPS Clustering Plus supports active-active mode on the network so that both active servers can read and write identical files locally and maintain data consistency and coherency for a given file on two active servers. This represents a giant step forward. Clients access real-time information from the closest active server.
Multiple Location Support: TPS Clustering Plus works not only in local clustering environments, as most solutions do today, but over SAN, MAN, and WAN in multiple locations. This greatly improves business continuity under disaster conditions.
Scalability: Active and passive servers can be added on the fly or taken out for maintenance and service as needed without affecting data availability.
Data Integrity and Consistency Check: Data integrity and consistency checks are run whenever a server is brought into service, to ensure it contains information identical to its mirroring partner or partners.
Automatic Data Recovery and Synchronization: Data recovery is performed automatically when a server is brought back into service after a shutdown, whether expected or unexpected.
Many-to-Many Mirroring Support: Data can be replicated selectively in real time, based on significance. Mirroring is performed on any directory within a file system, and different directories can be mirrored to different file systems on different servers. Granularity can be adjusted at the directory level.

Product Specification

Hardware: Sun SPARC Systems

Software: Solaris 8 and Solaris 9

Architecture:

Active server: Sun SPARC System running Solaris 8 or 9

Passive server: Any UNIX or Linux systems that support NFSv2, v3

Client system: Any UNIX or Linux clients running NFS v2, NFS v3; Windows systems running CIFS

Configuration: Two servers are configured as an active-active or active-passive pair

Scalability: Three or more servers can be configured as a daisy-chain, e.g., active-passive/active-passive

New servers can be added on the fly

File System Support: UFS, NFS v2, NFS v3

File System Size: No set limit; file system sizes on replication servers can differ

Storage Requirement: Any disk or storage, such as RAID or JBOD

Network Connection: LAN: 10M, 100M, 1G Ethernet; SAN: Fiber Channel; WAN: DS3 or above