# 电子工程代写|计算机系统原理代写Principles of Computer Systems代考|CSSE7231

## 电子工程代写|计算机系统原理代写Principles of Computer Systems代考|The Network Time Protocol (NTP) Method

In this method, computers in the system are arranged in a varying layered graph structure. A graph has an n-layered structure $(n>1)$ if its set of vertices $V$ can be partitioned into a union of subsets numbered from 1 to $\mathrm{n}: \mathrm{V}=\mathrm{V}{1} \cup \mathrm{V}{2} \cup \cdots \cup$ $\mathrm{V}{\mathrm{n}}, \mathrm{V}{i} \cap \mathrm{V}{j}=\varnothing(i, j=1, \ldots, \mathrm{n}, \mathrm{i} \neq \mathrm{j})$, called layers, so that edges may connect vertices from the neighbouring layers $\mathrm{V}{i}, \mathrm{~V}_{i+1}$ only. The structure is called a $s y n$ chronization subnet. Computers in layer 1 receive directly signals of UTC from a time service and are mutually synchronized. Computers of the $i+1(i<n)$ layer are being synchronized by computers of the layer $i$ by means of a certain chosen method, e.g. by the Cristian’s method, extended to a group of synchronizing computers. Therefore computers of the layer $i+1$ play a part of clients of computers of the layer $i$. That is why all computers in this method are called servers. Synchronization is all the more exact in layers closer to the time service, i.e. those with low numbers. Apart from their function as synchronisers and clients, the computers may intercommunicate regardless of their location in the graph layered structure. Communication links not intended for synchronization, are used for reconfiguration of the graph structure, when a certain computer gets faulty and is removed from the synchronization subnet, or crosses from one layer to another when a disconnection of synchronizing links with it occured. So, the graph structure may fluctuate: the synchronization subnet is subject to reconfiguration. An exemplary synchronization subnet in the course of working is depicted in Table 4.2. Computer $\mathrm{L}$ has passed from layer 3 to layer 2 and computer $\mathrm{D}$ from layer 2 to layer 3. Note that the synchronization subnet, for $\mathrm{n}=2$ is a bipartite graph, which, in this case, is the communication structure in the Cristian’s method but with more than one time server. The synchronization subnet is a mechanism for time management in various kind of networks. It became a basis for standard protocol of clocks in various kind of networks. It became a basis for standard protocol of clocks synchronization in the Internet (Network Time Protocol 1989). It takes into account delays in large networks to milliseconds, whereas in local networks – to fractions of milliseconds. Because of multiplicity of servers and links with clients, the NTP method enjoys a good degree of fault resistance caused by computers’ failures and loss of packets in transmission channels. For better efficiency, a simplified version of the NTP protocol (abbreviated to SNTP) resigns from some algorithms of the full protocol, not necessarily required hy some distrihuted systems.

## 电子工程代写|计算机系统原理代写Principles of Computer Systems代考|Distributed Mutual Exclusion Without External

As in the former methods with supervisory server and token-ring, it is assumed that there is one protective zone (extension to several of them is straightforward) and during its execution no failure of the system occurs. Processes are cooperating by broadcast of messages among themselves, taking advantage of global timestamps in this cooperation. Figure $4.4$ depicts structure of a system with four computers, arbitrarily numbered, that compete for entrance to the protective zone. Similarly to the methods presented in Sects. $3.4$ and 3.5, the computers send, ,request”, , permission” and ,refuse” messages in the course of the competition, but between themselves, not via any supervisory server.

These messages (of the form of system procedures) are interpreted as follows:

• Process which requests the protecting zone broadcasts the ,request” message with its current global timestamp as parameter, stores it in its register , global timestump of process which has sent request” and waits for replies from receivers. After having received all the replies, increases the content of its register „,number of granted permissions” by number of replies, permission”, and if among them is the ,refuse” message, passes to the waiting state. The ,refuse” message, the process receives from process executing protective zone and from processes being in the waiting state, whose global timestamps are smaller (according to the $\leqslant$ order) than its own. If the content of register „number of granted permissions” becomes by one smaller than the number of all processes, the process enters the protective zone.

## 电子工程代写|计算机系统原理代写Principles of Computer Systems代考|Distributed Mutual Exclusion Without External

• 请求保护区的进程以当前全局时间戳为参数广播“请求”消息，将其存储在其寄存器“已发送请求的进程的全局时间戳”中，并等待接收者的回复。在收到所有回复后，将其寄存器“，授予权限的数量”的内容增加回复数，权限，如果其中有，拒绝消息，则进入等待状态。,refuse 消息，进程从执行保护区的进程和处于等待状态的进程接收，其全局时间戳较小（根据⩽order) 比它自己的。如果寄存器“授予权限的数量”的内容比所有进程的数量小1，则该进程进入保护区。

myassignments-help数学代考价格说明

1、客户需提供物理代考的网址，相关账户，以及课程名称，Textbook等相关资料~客服会根据作业数量和持续时间给您定价~使收费透明，让您清楚的知道您的钱花在什么地方。

2、数学代写一般每篇报价约为600—1000rmb，费用根据持续时间、周作业量、成绩要求有所浮动(持续时间越长约便宜、周作业量越多约贵、成绩要求越高越贵)，报价后价格觉得合适，可以先付一周的款，我们帮你试做，满意后再继续，遇到Fail全额退款。

3、myassignments-help公司所有MATH作业代写服务支持付半款，全款，周付款，周付款一方面方便大家查阅自己的分数，一方面也方便大家资金周转，注意:每周固定周一时先预付下周的定金，不付定金不予继续做。物理代写一次性付清打9.5折。

Math作业代写、数学代写常见问题

myassignments-help擅长领域包含但不是全部: