Network Technology and Technology revolution weekend Caprice

Ignorance of the best not to squirt dung public Friday back home from work to take the subway, since the wife went to the field for a few days to open a meeting, I weekend to do full-time dad, I think it will be exciting. Can be a little alone with my family, deepening the feelings about his daughter, so humming songs. Looking at the same time "hacker big exposure". Occasionally glanced at the beauty beside it ...
A voice immediately let me have a want to hit the impulse, so next to the beauty in my eyes turned into the sky smoke.

The sound came from behind me. Through glass reflection can be seen is a very ugly young female program ape, height of 150cm less than. Weight roughly greater than or equal to 130 pounds, shouting: The network is not studious, just to understand the general principle. Can pass, now who also engage in network Ah, not a few years. All the networks have to go home!
It's disgusting, her misfortune is that she met me. Her luck is that we are in a public place. She and a fellow man talk about the TMD, as if it is to do storage, but also very low ground. Quite the bottom. Assuming it's not her ignorance, how can you be so rude and know what a storage network is? Alas. Today's remote memory access to the use of TCP/IP, the network almost penetrated into whatever field, assuming that some people still think that the network is ping or traceroute so simple, directly resign themselves. If you do not resign, then open it straight away!

But I very quickly forget such an unintentional attack on their professional quality, after all, I do not know that I am engaged in the network, and then. I am not a professional network, just out of the love of network technology has always been as their own career. This confirms the sentence. You can insult me, but you can't insult my career.

Best Practices for learning Network layering models and network principles received an e-mail, asked me how to learn the network, and too lazy to read, I would like to learn from the practice, I can not answer to say that when the network management, because he may not be able to apply to the ideal post. Can learn practical things job recruitment is a certain strength, Internet cafes network management is expected to, but do not learn anything. The same reason, I can not let others to apply for the application of ape. So what? can only toss themselves. Think high-speed is not impossible, the key to see what you toss, so write down the following paragraph, hope to be helpful to others.

Suppose you want to learn the Internet, and want to thoroughly understand it, then temporarily throw away those TCP/IP books, and so on understand the principle and then look back, and do not plunge into the Linux kernel stack, it is just an implementation. And the implementation is very not completely.

The best way is to learn the mobile communication network, while the mobile core network is not fully IP before, hurriedly seize this good time.
Undoubtedly. Mobile communication network is complex, such as Umts,lte, and so on, the terminology is also very many, standardization organizations and miscellaneous, a variety of technical aspects or legal disputes continue, but the vitality of unlimited. You know, the mobile communication network is the first and only one that is compatible with super multiple types of networks. It can be connected to the PSTN and can also be connected to a pure TCP/IP network. From 30 ago, has been developed to today. Despite the gradual convergence of technology, some non-technical factors have slowed the process. Learning this network, you will understand almost all the knowledge about the network, but also understand the general rules of the development of the network.
The complexity of the mobile communication network is obvious, but the reason we find it complex is very subtle. Since all along. The network courses taught by the University are based on TCP/IP, the link layer is almost not involved, non-IP network layer protocol almost does not involve, IP routing outside the network addressing approach almost does not involve, almost all are speaking IP address, IP routing, TCP principle. The UDP principle, as well as some common application protocols, for example HTTP,FTP. ... Let's say you've been involved in some vendor training. For example, H3c,cisco's advanced project teacher training may be exposed to a certain range of ATM. X.25,PSTN, but this kind of training is very expensive. And almost none of them are for research and development. Unless you go to companies that actually develop ATM switching devices or develop mobile network devices. Otherwise, almost all of you will be exposed to TCP/IP. So, in the face of a new thing, of course, you will find it complex. Everything except TCP/IP is new to most people.
TCP/IP is a blueprint. But it was not a reality, though it almost defeated all its rivals in the 1980s. But that's just theoretical. The first-place transfer process continues until today, in theory, more than 40% of manufacturers claim to support the TCP/IP stack, which means that TCP/IP wins, but the reality is not TCP/IP a single big, for the compatibility of the maintenance is blocked TCP/IP dominate the world factor. Although vendors claim to support TCP/IP, it does not mean that they do not support other stacks. Look at Shehiren's textbook, starting with the 2nd edition. Gradually removed the addition of TCP/IP, which also shows that the success of TCP/IP is a gradual process, the name of LTE can be really good.

For the encapsulation of the Protocol, most people learn to take care of the layer. Transport layer, IP layer, Ethernet layer. The physical layer is encapsulated in this order, assuming that some other way of encapsulation is also nothing more than IPSec Esp/ah,pppoe. SSL/TLS. Gre,ipip, even for people like me, OpenVPN in TCP/IP networks, tunnels are an uncommon way. However in the real mobile network. Tunnels are a super-common way, in general, when you visit Google home on your mobile phone, you create a tunnel between your phone and your mobile core network. The tunnel number changes with the mobile phone, inside the tunnel is the real phone to the Google home TCP/IP data, but the tunnel use IP protocol? The answer is to be able to use or can not use the IP with what? Very much. For example, ATM. Let's say you know that it's not practical to call and send text messages at all. That means you're on a level, and they're still going to be circuit-switched, and you won't see any TCP/IP information when the data is starting from the phone, so don't expect a packet-swapping tool like tcpdump to grab it. Further, if you can know that even the circuit exchange data can be encapsulated in the RAN and core network in the TCP/IP protocol family, then your hierarchy has been improved again, TCP/IP can be used as a bearer network, but also as a carrier, which is the advantages of protocol layering. X over Y,x. Y can be whatever the protocol, no matter what the protocol, is the essence. Further, suppose you can know very much about mobile data, whether it is a circuit-switched domain or a packet-switched domain, mostly using UDP encapsulation and not using TCP. You are already a prospective expert on the Internet. Since you know nothing about the nature of the layered model. Also know the nesting standard principle of each layer mechanism ... Even the TCP/IP family has borrowed a lot from the mobile communication network. The mobile IP mechanism is one of them.
Suppose you ask me what level I am actually half bottle level. At least better than the empty bottle, I think the bottle is not what derogatory ... Well, how do I know that? First, I only looked at a little bit of TCP/IP in college and then took the H3C training, when Cisco training was too expensive. Can't afford to play (still can't play Now). There is no burden; second, I study the internet not to get a high salary, because I know that my education experience, can not get any high salary, just to find a job. Without such utilitarian, learning is not the focus, because a lot of people are generally recognized as the key or fashionable things assumed to understand, you can get high salary; third, and the previous point similar, I do not participate in any test, the first to participate in the H3C training exam is a very large part of the teacher, do not participate in the test will not do the problem You don't have to memorize the boring stuff, of course. The cost of not getting a certificate to get a high salary is that I learned a lot of really useful things.
Imagine the router shunt mechanism the Bourg paradox of urban traffic network exists in the Internet, which refers to the Internet based on IP. As the current IP Internet is based on the shortest path routing data, although there are also very many policy routes inside. But there is no fundamental change in the state of traffic congestion. About the Bourg paradox and Nash equilibrium Point can be involved in the wiki.

In urban traffic networks, routing decision makers are drivers. While on the Internet, routing decision makers are routers. Whatever it is, it is addressed in accordance with a selfish routing strategy, and in general it is a third-party diversion.
In a closed selfish routing game system, it is generally possible to reach a consensus of maximizing common interests. This consensus is the so-called equilibrium point.

Traffic networks and interconnection networks now rely on such a balance point for execution without crashing, but at the cost of having to face a sudden congestion problem.

The

assumes a slight change to the router's algorithm. Routing packets no longer follow the shortest path. Is it better to shunt the packets according to the cost weights of all the accessible paths? It's really hard to say that this is the game of centralized control and distributed control. Assumptions are based on the shunt algorithm. A lot of data go is not the optimal path, but the advantage is to avoid congestion, those who do not take the optimal path of the cost of the packet is worth it? Since the network has been executing for so many years. Let's keep its distributed control going.
       and traffic regulation, the more congestion and more governance, the more control congestion, congestion tax. Limit the number of lines. Just because you don't have traffic control, there is no way to manage congestion. The only way is to leave it alone. Endure a threshold, the driver stopped on the road can not move, the situation should be eased.

It's just that. The current Internet routing system is nearly as close as the traffic network, since most routers weigh path metrics on a static weighted basis. There are delays and throughput for the participation factor. Transmission medium, physical distance, on the BGP may also have political factors, etc., it can not reflect the real-time status of the path, but, because the traffic aggregation of the shortest path algorithm in the selfish routing continues to function, the measurement will inevitably change with the traffic, assuming to support real-time measurement monitoring, it is necessary to Exchange routing protocol data Frequently, Then it converges according to the new calculation results. Such a control channel data volume is considerable, and more unfortunately, it will be fed back into the existing computing process, because the Internet routing protocol is implemented in-band.
So. Although there is not necessarily a need for centralized control, there are two points that I think are necessary. The first is that the routing protocol moves out of the band, like an early traffic network, that can be measured by helicopters. Other than that. The self-routing mechanism that implements some kind of packet may be able to slow down the constant computation pressure on the router's control plane.
On the technology revolution the Internet as the third technological revolution is a little exaggerated, it was not previously said that the PC revolution is No.3 it? In fact, the PC. The internet. are not worthy of the technical revolution of this title! The most fundamental thing that the technological revolution refers to is the source of human life, the first is the agricultural revolution, conquered the settlement of human beings, people no longer drift. It promotes the accumulation of knowledge and technology objectively, and the second time is the Industrial revolution. Overcome the human power problem, and further let mankind start drifting again, people's homes again back to the state of continuous movement, in fact, the accumulation of knowledge caused by the agricultural Revolution created the industrial Revolution, in fact, electricity, PC and the Internet is only the continuation of the Industrial Revolution, Bring geographically dispersed people together via electricity-pc-the internet. To make people physically move in a state of drift, but in a logically settled state; the third may be the genetic revolution, which completely blurs the line between machines and life, and this is the idea in the bestseller "Runaway".
The assumption is that the internet is a revolution. So the train, the car. Phone calls and so on only to create a technology, can be called the Revolution!

In fact, a technology is seen as a revolution many others are propaganda. And also. Say what jobs and the like changed the world, but also the nonsense, you have changed the world, after all, more than you so personal! "Changing the World" is the most troublesome argument I have! There are many people who are corporate or personal propaganda! We can talk about changing the world with only two people, Adam and his horse!

Copyright notice: This article blog original articles, blogs, without consent, may not be reproduced.

Network Technology and Technology revolution weekend Caprice