Ensure The Server System Is Robust Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Network operation and internet is the main components for nowadays company and enterprise. Moreover, Server system and database which deal with all of the data delivery process for the company are playing core roles of the network. Any human errors, hard disk damage, computer viruses, even the natural disaster will cause unpredictable loss.

The failure of web server and database will not only cause the loss of system files, transactions, customer information, financial problems and etc., but also it bring disastrous consequences for business, such as business bankrupt , close down. Therefore, effectively protecting the existing data and making the stable operation of the server system become every important issue. In order to ensure the server system is robust and minimizes the business loss, we need to restructure the system and enhance human resource of the company.

The remainder of this paper is organized as follow. Section 2 introduces background of the company base on the lecture notes. Section 3 relates the similar problems are discussed on other people's research. Section 4 details contingency plan and how to ensure the server system is robust. Section 5 concludes the discussing and future research.

Traditionally, the company utilized broadband internet access technology which includes wired and wireless connection. Every staff can access the internet and company's server though their own computers. The company is running various servers for its business such as web server, mail server, and database server. Also, the company adopts many open source and freeware to build existing server system in order to achieve cost effective, such as SMTP server, apache and etc.

Apache as an open-source is applied for web server in the company. The web server of the company provides the main website, web forum and also connects the database server in order to store and search the information. In addition, due to the huge number of information and web forum on web server, the company uses one domain names and three web servers. MySQL is selected for database server of the company as a back end service to analyze and store the data, because MySQL is also utilized in many large enterprises such as Wikipedia, Google and Facebook (1). Furthermore, the most popular freeware is adopted by the company for email system in order to provide cost effectiveness and stable service, such as SMTP, POP3 and IMAP server.

3. Related work

We found some similar works to discuss how to make the system more robust. UCSD Server Room Standards (3) present that the server location and server room infrastructure elements must be proper. For instance, backup power supply, separate server room (we discuss in section 4), extra cooling redundancy and etc. must be considered in the server room construction. Moreover, the following section discuss that we use extra switch and dial up system to build spare internet connection, however, Routing Architectures by Sam Halabi (4) introduced a tool called BGP (Border Gateway Protocol) can achieve internet connection redundancy. The basic concept of BGP is that providing best path to resource by choosing ISP to achieve network optimization and redundancy. Furthermore, windows server 2003 introduced one of server redundancy technology called clustering by utilizing fault tolerance concept. Cluster provide more than one application server to operate only one application. Therefore, if one server is out of service, the others can take over the role and run the application. (5)

Manpower always is important issue for us. Many huge enterprises discussed how to manage IT human resource. They trained the most significant IT people in the core position in the company, also IT department manager re-evaluate all of IT staffs corporate with human resource department, even a new job position appeared, called human resource IT manger.(7) Therefore, we also will discuss manpower redundancy in the company in next section.

4. Solution details

Not only the company can adopt clustering technology, BGP and other redundancy as we explained in section 3, but also in this section, we will restructure the company server system base on redundancy, Server room configurations and human resource, also provide contingency plan against system failure problems arise.

4.1 Tolerant backbone

The fault tolerant backbone network will be adopted by our company in order to enhance the system. Inside of the company, the web server, mail server and database server will be located into three different places, and connect together with two channels. The additional switch is used as an emergence channel which provides lower speed of internet access. Therefore, suppose that the company has network connection problems, one channel is fail, the other one is still working. (Figure 1)

4.2 Keep dialup internet access

Not only broadband internet access (mix of wired and wireless connection) in each spot of the company, but also telephone wires in each office in the company. Therefore, the company will provide dialup access service to all the offices. If the broadband internet connection is failure, at least we can connect each individual computer to the company's server internally through local telephone wires in order to continue to run the business.

4.3 Proxy

Huge number of user contains internal and external users access the company the website everyday. Local proxy is set up by the company to increase the internet access speed and share the burden of the web server. When the company's staff and outside user access the main web page of the company, they will access local proxy first, which has stored the requested web page, and then send the page back to the browsers. Therefore, Web server will be more stable, because it handles the fewer requests. (Figure 2)

4.4 Round-robin domain name server

The company can use both http redirect technology and round-robin domain name server. Due to one domain name and three web server in the company, DNS server will be add into the system and set one web server as main server. First, user access DNS server, and the browser get feed back form DNS server. And then, the browser can connect the main server, according to the diagram, the redirect message contains the IP address of the selected web server. Finally, the browser can connect web server 1 or web server 2 bases on round robin method. Hence, the 1st user request web server will be send to web server 1 and the 2nd applicant will be assigned to the web server 2 and etc.

4.5 Equipment redundancy

The company can use limited amount of equipment redundancy/machine redundancy to improve the effectiveness of mail server. The mail server in the company will be divided into three levels servers. The first level server will filter spam mail and scan virus, the second level server can dispatch emails to mail servers and the third level server contains SMTP/POP3/IMAP4 servers and distribute different server groups. The staffs will be separated into different groups to access the mail servers.

We will use 4 computers and set up each computer as standby redundancy machine for each level of servers. If web mail server 1 will be utilized, the other web mail server is redundant. (Figure 4)

4.6 Data redundancy

For the purpose of keeping data for long time and provide better reliability, the system will not just record each data in two different hard disks at the same time, also provide daily backup, one weekly backup copy and one month backup system.

We will use the highest level of RAID method (RAID 1) to ensure users data maximum availability and recoverability. The system will write 100% of user's data in hard drive, and automatically copy the data to another hard drive. (2) In addition, data as the most significant component for the company will be maintained and managed by data management professionals.

4.7 Manpower redundancy

Frist, the company will train the every single IT staffs, so that each employee of the IT department will understand the core technology. If the company has manpower problem, we can do find suitable candidates by using internal search. Second, we can hire some staffs as part time. One job may have different part time employees. So suppose one person quit the job, the other one still can work for us. In addition, the company will maintain relationships from other partner company, if we have manpower redundancy problem, we also can borrow some people as outside recourse form Partner Company.

4.8 server room configuration

Not only Our Server room locates near the IT manager's office, also water cooling system, air-conditions, fire extinguisher, raised floor and sealed windows and extra electricity can be provided. First, water cooling system will be built under ground to take away heat, also redundant air conditioner can work in any emergency happened. Second, we will use gas-based fire extinguisher instead of water fire extinguisher. Third, sealed windows and raised floor can be provided to prevent water from outside. The last, we adopt UPS system as a redundant electricity supply.

5. Discussions and Conclusion

As we discussed above, both technologies and human resource can be adopted to improve the server system. BGP, cluster, equipment redundancy, data redundancy, internet connection redundancy, manpower redundancy, and server room configurations are main method to ensure the server system is robust. However, there are still some limitations about our solutions. Some major equipment as single access point of the system may fail; also high cost cannot be avoided, even security issue, power supply and Natural disasters problems.

As for future research, in order to provide more stable server system, our key focus is about Cloud Computing technologies. With rapid development of internet, the company will use Cloud Computing technology to ensure the server system is robust in the future. Currently, the company still concern about security issues and risks. With Cloud Computing, "cloud" will be calculated for storage and help us to save the data instead of database server. Each group of computers on the "Cloud" includes hundreds of thousands or even millions of computers. (6) Hence, user can access the database anywhere. Moreover, cloud computing also can improve business performance by reduce the cost for company, better virus protection, and stable process. Therefore, with help of "Cloud", we do not have to worry about any data loss in the system. People can predict that in the next decades, Cloud Computing technologies will ensure server system is robust in the company.