Mirroring as a Fault Tolerant System - Essay Example

Mirroring as a Fault Tolerant System September 20, Introduction Database shadowing and mirroring is a popular option that is used for fault tolerant systems. This supports a database by creating an immediate backup copy on a second drive. The mirroring effect begins with an initial drive that backups data through a reflective drive. The advantage to this is one that doesn’t require extra backup software and controls when one is working through the drive. More important, the bandwidth is increased instead of decreased from the storage area as the data stored provides flexibility of use within a hardware system. For those that want to avoid immediate repairs or computer crashes, is the ability to use this alternative as a simplistic way of retrieving data. Understanding the different components as a part of shadowing and mirroring as well as how it can be used to assist in databases can provide individuals with more alternatives for building the database security needed while creating higher levels of fault tolerance. The need to examine the benefits and disadvantages of the mirroring system is also required. This provides an evolution toward the main approach of mirroring and shadowing for more advanced technology that is currently being built for end users. Technologies with Types of Fault – Tolerant Systems Several types of technologies are available and used for the shadowing and mirroring fault – tolerance system. Each of these can work together to provide complete functioning within the system while allowing the backup system created to work more efficiently. The first technology used is from physical aspects of the system. The mirroring built through a hardware system is placed into the main drive of the computer. This divides by a coupled multi – processor with modules that carry dual paths. The cross - section and double paths as a part of the mirroring then create a way for data to move from one component to another. These interface with dual ported controllers and a disc subsystem that allows the data controlled with where it goes and how the information can mirror. The storage then takes place in the disc subsystem to provide complete backup to the system. Newer physical properties built from remote servers are also available. These specific options allow end users to immediately back-up data through an online portal connected through a remote server. The information moves into a storage space by connecting from one network to the computer so one can save all applicable information. The main hardware of this system built into a server provides a hosting company who has larger backup space on the computer (Bartlett et al, 1986). The second aspect linked to the fault system creates a base on the network built from the cross sections in the hardware. Processors are integrated into the system and maintain a high level of data sharing. The most common types of network integration systems are Systems Network Architecture, (SNA) and Open Systems Interconnection (OSI) as well as a proprietary network. These are built on top of the database to exchange information between the two components and to store information needed in each of the areas. The network for other components connects to a remote area so end users can save data through an online portal. The network connects directly to the database, which builds as a relational system. The relational database is then able to distribute the data between the two main network protocols and links and provides both distributed data and transactions between both sides of the wires. The relational database defines the wires that connect both of the drives together. As each work together, they can save and restore information in real time. The database is the main component of storing and exchanging information, there is no need to have software on the database, which provides easier management and easier implementation of fault – tolerance within a hardware system (Bartlett et al, 1986). The components as a part of the mirrored or shadowed database create the complete system that is used. This comes from hardware implementation integrated into the overall system. As the transactions and data are received throughout the system, it can move through the cross wires, data protocols. and other applications on the computer. The system includes a fail fast implementation, which automatically detects a fault and stops the main storage and moves instantly into the backup storage. For those in need of using the hardware system is the ability to continue working on the device without having to stop and repair the system. Many also use a real-time system to ensure that all data integrates and saves on both sides so there is no loss of materials and data. The dual path system created is then able to stop failures to once every 10,000,000 hours or once in 1000 years (Bartlett et al, 1986). The newer and remote mirroring systems use this same application; however, the integration of the data moves into a remote processor, as opposed to a system integrated into the hardware. The same basic system is set through the server, which can keep more of the information safe and secure as a backup through the system. Key Vendors and Cost Those that decide to use mirroring as the main component will need to contact a vendor that can host a hardware system. Unix, Linux and Microsoft are the main operating systems that support mirroring and shadowing. Most of the requirements placed in the hardware ensure that the operating system first matches with different needs. The suppliers that are available will then directly link to these suppliers as technicians that specialize in mirroring backup plans. Tandem is also a leading provider in offering solutions for the operating systems; however, the mirroring process integrated into the system and doesn’t come through extra hosting provided by the operating system. Sun, Solaris and other major computer providers have also provided different computers that carry specialized computers that have this fault tolerance alternative. Newer developments have begun to implement the hardware system remotely, such as Seneca, which provides more advanced technologies that can easily integrate into the main operating system. The price considerations for this will include the basic price of the computer and the hardware that is already integrated into the system with ranges starting at $550 a computer and moving into a higher end set of $2000 per computer. If an end user decides to use the remote mirroring process, such as Seneca, then the cost is $5 - $10 per month for the data storage use. More storage needed creates price differentiation because of the need to use more bandwidth (Leider, 2000). Advantages and Disadvantages to Safeguard Information Several advantages for computer users are seen through the use of mirroring. The first base is on the ability to have synchronized copies of work. Other versions of fault tolerant systems are not able to update information in real-time and often lead to different effects with the network that don’t provide all of the needed data in case of loss. The mirroring process can update and synchronize quickly and efficiently with the information that is being used. The second advantage is that there is not the need to stop the process of working if the data area shuts down or lost. For those who need to have easy access to their computer is the ability to use this fault tolerant system with the expectation that it will continue to provide abilities to work. Because of this, there is a lower expectancy of operation system crashes and other common problems among network protocols. The third advantage is that the system uses little to no bandwidth. Once the data transfers and is stored, it will also increase the I/O bandwidth, which allows the computer to function quickly and with ease, even if there are large amounts of data stored (Ji et al, 2003). The other benefits of the fault tolerance system used are based on the capabilities of the drive to have motor and drive topologies which are more effective within the system. This includes two main controllers that function as motors which connect to the main network of the system. The discs are then able to store information based on the consistency of the motors, which drive information between the two main networks. Instead of the computer accessing a different area, it can instantly place the information into a remote area without the complications and difficulties of another system that connects to the drive. If the fault tolerance system were going to use a different system, it would require software and extra components of continuing with functioning. The motor and drive topology exceeds beyond this and instead allows the main functions to come from the remote network or hardware system. This stops extra functions from being necessary to recover faults within the system (Ertrugal, et al, 2002). Even though there are several advantages to the main system, there are also specific problems and disadvantages that occur with the mirroring or shadowing effect. The first base is on how other technologies are continuing to grow and change within the industry. As this occurs, there are more complex features and levels of data that are used. The meantime to failure that happens through the main systems is becoming shorter and the ability to save as much data is also becoming more complex. For mirroring to continue to be effective, is also the need to continue to move into the complexities with technology. Many believe that the performance of fault tolerance systems won’t be able to survive beyond the node failures and will begin to abort or dysfunction. The mirroring system will then become unreliable beyond the two storage systems that can comply with the data that is stored. There is only one cross section for the data to store which can cause complexities and may lead to ultimate failure in the system or among the data storage that is on the system (Chen et al, 2005). The second set of components related to the failure of the systems is from the recovery part of the system. Mirroring and shadowing promises to save data at a real-time through two cross wires so individuals can continue working on an operating system. However, there is not the ability to recover the first cross wire and storage area. There are also complications with finding the status of the fault recovery system after it is restored to see if it can continue functioning through the same means. The result is that the communication and messaging may change after recovery and may limit the ability to store and save the data and information that is a part of the operating system. As the system continues to be used, it changes the abilities for the computer to continue functioning at the same level and will cause limitations with the backup and storage that is a part of fault recovery. Like other systems, the mirroring effect will need maintenance, looked into and processes for alternative recovery should be implemented to ensure that there is complete recovery of the computer. More important, checkpoint and round off errors can continue to limit this if there is maintenance, meaning that the system can continue to be affected through maintenance and alternate recoveries. The mirroring effect can provide high end options to continue operating but there are still more limitations with the process if a fault does occur (Chen et al, 2005). Conclusion The fault recovery systems that are available for operating systems can provide more options for those in need of saving and storing data. A popular method for retrieving data is through the mirroring or shadowing effect. This can assist with the recovery through a backup system that automatically takes effect in case of a fault. More important, this can provide more bandwidth and options for continual functioning ifs something happens to an operating system. Even though this is currently one of the most effective mediums for data recovery, there are several faults of the system, specifically with the changes in technology that are occurring. Maintenance, being able to restore faulty mirroring processes and changes in the process used for mirroring are all contributing to the disadvantages of the system. For the procedure to continue to work effectively is the need for operating system vendors to relook options for fault recovery systems while allowing this option to evolve into more enhanced abilities to save data. References Bartlett, Joel, Jim Gray, Bob Horst. (1986). “Fault Tolerance in Tandem Computer Systems.” Tandem. Chen, Zizhong, Graham Fagg, Edgar Gabriel, Julien Langou, Thara Angskun, George Bosilca, Jack Dongarra. (2005). “Fault Tolerant High Performance Computing by a Coding Approach.” Principles and Practice of Parallel Programming (10). Ertugrul, N, W Soong, G Dostal, D Saxon. (2002). “Fault Tolerant Motor Drive System with Redundancy for Critical Applications.” Power Electronics Specialists. Ji, Minwen, Alistair Veitch, John Wilkes. (2003). “Seneca: Remote Mirroring Done Write.” Usenix Events. Leider, Joel. (2000). “Beyond Raid 5: Mirroring Your Way to Fault – Tolerant Storage.” Winchester Systems. Read More