360 Systems DigiCart/E - Manual (Page 16)

14
• Installing an Ethernet Audio Network
While DigiCart/E is robust, it is not designed to be fault tolerant. In mission critical applications, two
DigiCart/E recorders should be installed so that manual switch-over can be accomplished quickly.
Servers and other network equipment are available with many options that enhance relability or provide fault
tolerance. Fault tolerant design begins with redundancy and utilizes physical separations to minimize and
manage the risk of failure. Consideration should be given to redundant power systems, servers and other
network routing equipment, signal cabling, and data security. 360 Systems is available to consult with you
during the system design phase to help identify and achieve your system requirements.
Power Systems
Power failure to the entire system or to one part of the system can be one of the most devastating
problems in network security. Fortunately, many solutions are available. Power can be sourced from
more than one service; e.g. an electric utility and local uninterruptible power supply (UPS). Servers can
be installed with redundant power supplies, fed by separate power sources.
However, there are some practical limits to the application of redundant power sources. For example,
dual power supplies in a server may not be enough to keep the server running in the event of another
type of internal fault, such as a processor or fan failure. Some servers are built with internal
redundancy to address these problems; others can be configured in clustering schemes to automatically
transfer control to a surviving server. DigiCart/E has only one power inlet and one power supply
system. If you are designing a critical application, you should consider employing redundant
equipment where single power supplies will not meet your need for operational security.
Servers
A wide variety of high-availability servers offer various levels of hardware and software redundancy.
Servers are available with redundant power supplies, processors, data busses, disks and I/O. Servers
may be clustered to provide “fail-over” protection for the network. Additionally, many server
manufacturers offer advanced “early warning” software that can monitor the hardware system for
impending failure.
Windows 2000 Server Advanced includes clustering software. When applied in multi-server systems,
this software can automatically detect the failure of a server and quickly redirect network traffic to a
surviving server, with users experiencing only a momentary pause in service.
RAID Arrays
RAID (Redundant Array of Independent Disks) is an acronym first used in a 1988 paper by Berkeley
researchers Patterson, Gibson and Katz. It described array configuration and applications for multiple
inexpensive hard disks, providing fault tolerance (redundancy) and improved access rates. RAID
technology does not prevent drive failures. However, RAID does provide insurance against disk drive
failures by enabling real-time data recovery during a drive failure without data loss.
RAID provides a method of accessing multiple individual disks as if the array were one larger disk,
spreading data access out over these multiple disks, thereby reducing the risk of losing all data if one
drive fails, and improving access time.
Typically RAID is used in large file servers, and transaction application servers where data accessibility
is critical and fault tolerance is required. Currently, RAID is also being used in desktop systems for
CAD, multimedia editing and playback where higher transfer rates are needed.