Network That Houses the Storage

¶ … network that houses the storage file server(s) keeping all files related to contract transactions between all divisions and their clients. Your task is to design only this segment of the network. Do not worry about the rest of the network and do not worry about database and web servers. You will not be using them in your designed solution. The main file storage server must reside with the headquarters. I would use a directory-based network where all components of the network are considered as objects. These include servers, workstations, peripherals, users, files, and so forth. The Network objects are stored in a directory that is replicated to the directory servers / domain controllers so as to protect for both performance control and for redundancy. This also protects against network failure.

The workstation will be bale to run different types of client information concurrently so that it can carry large loads.

I will use disk servers -- unusual on LANs to serve as backup for reliability, speed, and capacity. The server or directory security will provide access to the disk server

Network servers will act as a sort of file server. These manage requests by different users on a file-by-file basis. The files do not cross concurrently, but sometimes many users can simultaneously access the same file. Actually, this is no problem to set up for most current networks allow any server to act as this file server. I would prefer -- as you will see later -- a database server instead of the traditional central file server. But the network file servers can back up user files in the eventuality of disk trouble and the users can easily share and transfer files. This would be a valuable option to have and something that I would incorporate in my network.,

The Network Operating System (NOS) manages file security. The directory and file level specifies the files to which users are permitted access and the level of access that they can have to these files. Windows 2000, Linux, and UNIX, Novell NetWare, Windows NT and IBM's AIX are all examples of NOS. Aside from network security, NOS also provides other functions such as authorizing new users and network trouble shooting.

I would have a NOS that would provide direct support for disk fault tolerance . It would have disk mirroring, disk duplexing, and a software implementation of the Redundant Array of Inexpensive Disks (RAID) 5 standard. Disk mirroring provides automatic backup control, therefore users will not notice if one hard disk fails.

The Raid 5 system is particularly good since it uses three or more hard disks spreading the data across the disks. Parity information is simultaneously interspersed with the data so if any of the disk drives fail, disk errors are detected and corrected.

I would configure the LAN with a database server that would provide systems processing such as searching and sorting databases. This is more powerful than a file server since it also provides:

Less traffic

Enhanced security when locking fails speedier response time.

There are problems here in that the database server allows multiple users with access to the same file concurrently causing backup and lack of availability. This may be worked with.

I would also have to work with the security benefit of the database which needs intense scrutiny.

To create more speed and better load capacity, I may include an application server. This supports the application system and is specifically designed to meet the organization's needs. They can share the processing load between servers and computers and create a more fully distributed computing environment.

Since I would like to configure the whole with enhanced capacity for bandwidth, I would devise a LAN that would work with hubs and switches and would, in other words, be a star topology network. Examples are 10BaseT ethernet and 100VB AnyLAN networks. I would devise an intelligent hub that would have enhanced capacities for other features such as remote management capability.

Switches are an addition o hubs and would also provide this prospective network with support capacities such as multiple bandwidths, bridge functionality, and router functionality.

I will install a bridge that will filter data packets according to their various addresses. This will forward only the appropriate packages to the relent address. The packet moves on if it does not belong. This function seperates the various files and ensures that each receives its correct one without clogging up the system. It also provides data security so that each receives only that which he is permitted to have access to. Bridges help th traffic. They require, however, that network managers install and update their list of addresses.

Routers are also an inescapable piece of my network. The router does many of the above functions reading the destination address for information, examining the contents of each packet, and establishing the most efficient path. All of this is performed by routing protocols.

I will use dynamic routing whereby the router will be able to automatically reconfigure its information table. It will be able to recalculate the lowest-cost path in terms of load, line delay, or bandwidth. I may even have one that will be able to rebalance the traffic load across multiple links. I will use the Open Shortest Path First (OSPF) as routing protocol.

Routers are slower than bridge but provide better security. I may therefore use bridges and routers in conjunction. If I had to choose between one or the other, I would prefer the router even though its far more expensive and difficult to maintain. They also provide network management overhead.

I will want to prevent unauthorized entry of suspicious software. To that end, my solution will be to introduce software at a central site and distribute it throughout the network. In this way, all have access to the same software that is located at a central region.

To ensure safe file location, I will also keep files safe in a central location whilst directing files in use to the local workstation.