1-2 Paragraph Response
Question- "Optimizing Performance"
In order to ensure the fault tolerant network design with no point of failure, administrators can use different approaches. One of them is redundancy. A network without a single point of failure should consist of redundant network-infrastructure components like routers and switches. Moreover, these devices should be interconnected with redundant links. Redundant components, paths and services can also be included in the network topology to ensure its high availability. Another approach is replication. It is a way of providing a couple of identical instances of the one system or one subsystem, directing tasks or requests to all of them. The third approach is network load balancing. This approach provides a highly fault-tolerant environment without the need for specialized, shared hardware. In my opinion, the administrator should evaluate both requirements and resources in order to choose the best approach. However, the most effective approach is to ensure redundancy as enterprise network does not need especially stringent fail-over times.
The network can be optimized by tuning the network adapter. It also helps to increase network speed and performance. According to eActivity, it can be done through enabling offload features and enabling interrupt moderation. It is very beneficial to have offload features turned on, as some adapters usually even require enabling this feature. By interrupting moderation for CPU-bound workloads instead of the increased host CPU savings, your adaptor will avoid more interrupts and less latency. Windows Server 2012 supports both options, so users may consider these ways of tuning.
Response to Samantha Jones:
I agree with Samantha that you are trying to ensure that there is no downtime when designing a fault tolerant network. The approach of software redundancy at layer 3 of the OSI model is widely used and really helps to ensure the fault tolerant network design, however, I do not agree that there is no one single approach. In my opinion, the redundancy is the best option in such cases. Both strategies for tuning the network adapter are described briefly and help to find the most useful one. I agree that Windows Server 2012 supports all these features and can be tuned if needed.
Question- "Command Line Tools"
It is a very difficult task to choose between Windows servers, UNIX servers, or Citrix Servers. First of all, we do not have enough information about business needs and existing servers. Moreover, each server has own features that can cater to the different needs of each business. The right answer here will be choosing the better choice given your business needs. UNIX systems are more stable than Windows or Citrix Servers, as they can handle extremely high up-time. Windows and Citrix servers perform better in low-stress conditions, but UNIX are the best for the high load. Moreover, the Windows servers are considered to be the most expensive in terms of hosting. We can surely say that UNIX-based servers are more stable, perform faster and more compatible than Windows servers.
Personally, I would strongly recommend a UNIX infrastructure, but it depends on the needs of business and whether there are unique MS products that do not have open source equivalents.
Unfortunately, the administrator doesn't have a choice of which program to use when it comes to using the "ipconfig" and the "ifconfig" commands. The "Ifconfig" is used in the UNIX environment and the "Ipconfig" is used in the Microsoft Windows environment. It depends only on the type of operating systems that are used in the network.
Response to John King III:
I agree with John King III that this scenario does not provide us with enough information about the type of business, business needs or its existing type of "network servers". It is very hard to compare or contrast these serves without knowing the criteria that is important to this particular scenario. John briefly explains main advantages of all servers but does not come up with his own suggestion, even though UNIX servers can be a great choice here. I completely agree that administrator doesn't have a choice of which program to use when it comes to using the "ipconfig" and the "ifconfig" commands.
Responses and Answers CIS 111
Question - "Database Life Cycle"
According to our text, the Database Life Cycle includes different phases. The actual implementation of each phase will be changed according to the size of the related organization or other organizational factors. As an example, we can choose the Implementation phase.
This stage involves the construction of a database according to the specification of a logical schema. Usually, it includes the specification of security enforcement, external and storage schema, etc. The choice of DBMS, database tools and operating environment will influence the Implementation phase. Taking into account that the data should be added to the tables, different important aspects of corporate data management should be supported, so implementation require a specific and detailed knowledge of features and facilities. First of all, we need to match the requirements with the available implementing tools. In this case, the size of related organizations and other factors will highly influence the phase.
Based on the way of distribution of the organization and associated distributed database design, the additional flexing of the design will overcome any software or hardware limitations.
Question - "Transactions and Concurrency Control"
When the data is attempted to be modified by many people at the same time, a system of controls should be implemented in order to not adversely affect those modifications of another. It is concurrency control. The concurrency control consists of two classifications of the methods of establishing concurrency control. For our example, we will choose Pessimistic concurrency control. To prevent editing data in different ways by one user that affects other users, the system of locks is used. It means that when one of the users performs some action may lead a lock to be used; others cannot do any actions that may conflict with the locked one. This method is named pessimistic control because it is usually used in different environments where there is a high contention for data. Additionally, the cost of protecting data with locks is a way lower than the cost of rolling back transactions in case of concurrency conflicts taking place. The advantages of using this method are its easy implementation and guarantee that all changes to your database are made safely and consistently. The biggest disadvantage is that this method is not scalable.
In case of not having this method available, business operations would have to change the data locking to errors that roll back the transaction and start over. It is called optimistic concurrency control. While using this method, the environment will change to the one where there is low contention for data, and where cost of locking data when read is a way higher that the cost of occasionally rolling back transactions.
