Subtitles of the Movie

Another main concern for our hardware is the actual infrastructure of our network itself. We really need to be aware of everything that's going on in our system when it comes to optimizing our database itself. So, what is our infrastructure? Well, truly, it's our architecture, specifically the architecture in terms of the equipment itself and the connections between those pieces of equipment. That's really what makes up our network. Now, keep in mind in talking about this architecture that the bandwidth and the latency are 2 of the most significant factors. Keep in mind that bottlenecks really occur when bandwidth is used up so we want to constantly be tracking how much of that bandwidth we are using and what specific times of day become our own rush hours. Keep in mind that high latency is a result of our applications running in the background. Now, if you have a particular application that is going back and forth consistently, this is going to help create those bottlenecks themselves. That's why it's always so personal to our particular system. We need to know how the network is created in the background and what are our demands on that network itself. Now, keep in mind there are some factors that will affect our bandwidth and our latency. For instance, we have our actual physical network. Now, the physical network is all about the machines being connected. What is their physical topology within the office itself? Do keep in mind that the structure of the Internet Protocol is what allows for the different networks to be connected physically. However, where are they located within the office? Once we get beyond the physical network itself, then we have our logical network. Now, in a physical network, we can group all the computers together within one room. What if our rooms are separated? Here's where the Logical network comes into play. They will be able to connect different areas via separate IP networks. When the machines are just connected to the same backbone, which is really the wiring, we can have separate networks all combined together. Here's where it comes in to be Logical. Now keep in mind when we're dealing with the physical and the logical network that we have 3 components to think in terms of. Yes, there is where the computers are physically located in the office, but then there is also security concerns. There could be one machine in the middle of the room that we still don't have any access to. And don't underestimate the power of politics in the office. Many times, politics will dictate both the physical and logical layout of our network itself. Now, in our network infrastructure, we also need to talk about the aspect of performance. Performance is always of great interest. When you're looking at the performance, we really have to think in terms of what parts of the object do you really wanna monitor. They'll differ depending on the object itself. So, when we talk about the aspect of performance of interest, we're really talking in terms of what part of the infrastructure do I want to test- Do I wanna test latency? Do I wanna see if my physical topology is correct? In this instance, we also need to talk about that instance itself. The instance refers to the specific object that we're measuring. If we have multiple objects that are all within the same system, that instance will be the particular server. An example, in SQL, we refer to an instance as being the name of the SQL Server database engine. It's important to know how we break down the naming convention to find what you're looking for within your network itself. There are, of course, some tools that we can use for the infrastructure. The testing is all again the terminology used in the performance monitor. In the performance monitor, we can test the current bandwidth. This will tell us exactly how much bandwidth we have, we can even find out exactly how much bandwidth is being used. Our packets per second, how many packets can we transfer per second and also our bytes total per second. This will tell us exactly how many bytes we can use per second. Now, as before, it's all about comparisons. We need to capture some of these traces so that we can compare our logs against each other. It's good to have a baseline, good to have an average. All of our performance monitoring is all about knowing what is the optimal condition? What are we doing now? How do those 2 compare to each other? Now that we know some of the tests that we can run, some of the components that we're looking at, let's go ahead and actually see some examples of the tools that we can use.