Troubleshooting for network engineers or network operation and maintenance personnel or even a lot of people (out IT friends) are very important, so know every step of the process and the principles of a detailed process of troubleshooting network is very important, when the network appears when the problem, we can quickly troubleshoot a problem on your network to quickly solve network problems.
Network troubleshooting here to say, it should be suitable for most people have some understanding of the network, in fact, on the method of network troubleshooting, have most of the information on the web, but I personally feel that too much just to stay in the text on the level of some of the basic principles involved are not mentioned, so practicality is not great, here I want to write a technical illustrations and a certain network troubleshooting articles to everyone, so that we truly master a set of system network troubleshooting methods.
An idea of this article:
- First, a prerequisite for network troubleshooting
- Second, the basic idea of network troubleshooting
- Third, the network troubleshooting detailed steps
- Fourth, *** IMPORTANT
First, a prerequisite for network troubleshooting
Why prerequisites? Because here by troubleshooting the network and not just stay in a little using a command, but a systematic approach, without these conditions, I can not really say the following guarantee these you can understand and can apply to practice, we still look at these conditions based on it.
1. familiar with the OSI model and the TCP / IP protocol stack
This should be a network troubleshooting to know the most basic knowledge of.
Whether OSI seven-layer model or DoD model, is used to describe a process network traffic in order to allow us to send and receive data on the network have a general process. OSI seven-layer model corresponding protocol of the OSI model, DoD DoD model corresponding protocol model, which is the TCP / IP protocol stack.
TCP / IP protocol stack inside it is very important, such as DNS, TCP, UDP, IP, ICMP, ARP, the most basic should have to know it, we do not have like a "TCP / IP protocol stack," said was so fine, but at least, the basic function of these agreements is that we have to be aware of, I think one or two of these I can not say clearly, and then again he Baidu about it.
Learn OSI layer 2 network communication infrastructure and its corresponding
Like switches, three switches, routers, firewalls these basic network device should be some understanding, especially their corresponding levels and the role of OSI, such as standard switcher corresponding data link layer of the OSI model, it collision domains can be isolated, and may be isolated from the virtual broadcast domains by LAN technology (the VLAN), switcher forwarding data frames is achieved by the MAC address of MAC address; another example, a router, which corresponds to the network layer, provides routing address functions and so on.
Probably you do not need to know too so fine, but probably a process that I think it ought to be aware of.
3. understand the basic architecture of a network of small and medium enterprises
In fact, the above mentioned is to know how those network devices are generally connected together, if debugging in a real environment, we must be aware of the basic structure of a current network situation.
Brief mention here, the basic structure of most small and medium enterprise networks are like this: the access layer - aggregation layer - the core layer - net exports.
FIG give below:
If the network environment is huge, aggregation layer and core layer will have, like here, then it is not directly aggregation layer, but the idea is the same. In fact, I say the following troubleshooting instructions are directed to the user's PC, in fact, the data center network troubleshooting is similar. No matter how complex the network environment, in fact simplified down with this it is similar.
4. know the common network troubleshooting command
If you are using a Windows operating system, then the following commands is very important:
If you are using the Linux operating system, it also has a corresponding commands. In fact, the final analysis is to be familiar with the equipment you are using, of course, in the actual network troubleshooting, if only to test whether the normal network traffic, you can put yourself familiar with the operation of the equipment replaced.
5. well aware of an important principle of network troubleshooting
Network troubleshooting, network does is to determine what went wrong, so you have to know the trend of the data.
So this is an important principle: attention to the data.
The following talk about network troubleshooting idea is actually to track data through a step by step to reduce the network point of failure, therefore, always remember this very, very important! This is why you need to get on top of SMEs understand the basic architecture of the network .
Second, the basic idea of network troubleshooting
The basic idea is as follows, with which most people write online should be the same.
If (1) Check the physical link in question
(2) Check the local IP address, routing, DNS settings if there are problems
(3) patency test the gateway or router. The first test and then test the gateway routers, test stage by stage
(4) testing patency of the public network ip ping (usually referred to several external IP)
Patency (5) DNS tests can be directly ping the Web site address
The third section below, the basic idea for the above will be described in detail.
Third, the network troubleshooting detailed steps
In order to better tell the network troubleshooting process and ideas, suppose we have one of the following network environment:
(Note: Although it is assumed that, but in fact the network environment is set up up by bloggers GNS3 linkage virtual machine and the real network, it can be a real reference)
Here, we have to top this network environment for example, details of our network troubleshooting ideas, how to do each step, after each step, and why do we do to get what information will make a note.
If (1) Check the physical link in question
I personally think this is a step in doing network troubleshooting steps must do ***! Often heard friends say, not on the leadership of the computer network troubleshooting needs in the past, engage waited a long time, still can not find the problem, ** * when several despair, has found no network cable connected to the computer. This really is a tragedy, and wasted a lot of time is not to say that such a network troubleshooting ideas have always been wrong.
Perhaps because not everyone can go to the room to see the switch wiring situation, so this step, we focus on the scope of the investigation should be placed where, as shown in the following figure:
In this step, the following points need to pay attention to the investigation of:
1. Make sure the computer network card itself has no problem
2. Verify that the network cable connection there is no problem
3. The switch unit is connected (if you can go to see the room)
If the above troubleshooting these points there is no problem, then the problem is that other devices in the network environment. The scope of the investigation is relatively simple, because only the connection problems related to the physical link.
For this test, you can consider using the measuring line device, but a personal recommendation is to take a laptop configured properly to do the test is also not a bad idea.
(2) Check the local IP address, routing, DNS settings if there are problems
*** above steps, the investigation of the physical link is no problem, that is, after the computer access the Internet line, the computer has responded, and can be identified, but the network still could not come to this step, it should first be noted range on the computer is set up above.
This last step, our focus is:
1.IP address settings
If the method is used to automatically obtain DHCP, then just look at this time there is no open automatically obtain IP settings, and there is no open-related services on their own to set the machine;
如果用的是静态IP，那么就必须要注意IP地址的填写有没有错(一般网络管理人员给的)、IP地址的子网掩码有没有问题(这很重要，对于静态IP，很多人在这里设置错误，建议是，***把IP地址、VLSM这方面的知识学一下)。
一般可以用下面的命令查看：
相关的设置可以在自己的计算机环境上面查看了。
2.路由设置
对于服务器、PC，一般是指默认网关的设置了;对于路由器本身或三层交换机，那就是静态路由或动态路由的设置问题了。
3.DNS设置
主要是要确保所设置的DNS服务器地址到底有没有提供域名解析服务或者是否出现了故障，至于如何判断，后面会给出方法，这里关注的是，你得设置一个正确的DNS服务器地址或可以自动获取。在windows上面你可以通过下面的命令查看：
(3)测试网关或路由器的通畅情况。先测网关然后再测路由器，一级一级地测试
在上面的网络环境中，在网络通的情况下，我们在电脑上使用命令tracert -d命令，会得到下面的结果：
通过这个测试结果，我们可以清楚地知道电脑在访问互联网时，数据的走向情况：
根据这个数据走向，我们就可以得到一个重要的思路，就是根据数据走向来检测网络的通畅情况!因此，我们可以分两步：
1.先测试电脑到网关192.168.2.254的通畅情况
我们可以在自己的电脑上自己ping网关的地址，看是否有响应：
一般这样的判断方法是比较快的，但有时候，无论怎样ping都不能，那么则可能有以下的几种情况：
a.网关设备做了禁止ping的设置
b.网关接口或网关设备出现故障
对于a，一般很少会在这些设备在做ping的限制操作，因为目前很网络设备如路由器或交换机的监控，都是通过ping操作来完成的。ping通192.168.2.254网关后，再ping一下172.16.13.1以确认电脑到整个网关设备都没有问题。
对于ping不通的时候，我个人还建议在电脑上执行如下操作：
即查看电脑本身有没有获取到网关的MAC的地址，显然，如果没有网关的MAC地址，那也是不可能ping通网关的，在排除了前面电脑设置的问题后，你可以猜测是网关设备出了问题，这时就可以联系网络工程师对网关设备进行测试了。
2.测试到其它路由器的通畅情况
前面一步没有问题了，也就是电脑到网关通信正常了，再测试网关到出口路由器的通畅情况：
这里，我们使用tracert -d命令就可以了：
当然，如果发现不通，那么则可能是下面的情况：
a.网关设备与路由器之间的物理链路问题
b.网关设备与路由器之间的设置问题，比如路由协议、接口配置之类的
出现上面的情况，那就是网络工程师的问题了，当然，如果你是网络工程师，应该要马上查看一下设备的状态，看是不是设备哪里出现问题了。
上面的步骤完成了，假设你的出口路由器设置是没有问题的，比如NAT与默认路由等的设置，那么我们大致可以知道，内网的一个基本通信是正常的(至少你的电脑和出口路由器的通信没问题)，我们就要看看电脑到底能不能访问互联网了。
(4)测试ping公网ip的通畅情况(平时要记几个外部IP)
来到这一步的时候，就说明前面三步是没有问题的，也就是说，本地局域网络的通信是正常的，这时要做的就是判断本地局域网络与外网(公网)之间的通信有没有问题了：
这里采取的是直接ping公网地址的方法，是为了排除DNS的影响(万一你的DNS设置又有问题)，至于要ping什么样的公网地址，个人建议是，可以ping一些没有禁止ping的公共DNS服务器地址，比如114.114.114.114和8.8.8.8的：
这 样之后，基本上就可以确定网络是没有问题的了。当然，这里并没有提到出口放置防火墙的情况，实际上，思路是一致的，但是，你需要考虑的是，你的访问数据有 没有被防火墙给过滤掉，是数据出去的时候过滤了，还是数据回来的时候过滤了?由于还要涉及到防火墙的设置，这里就不再提及了，只是仍要注意这一点就是了。
当然，也不排除是ISP那边的问题了，打电话去问问吧。
(5)测试DNS的通畅情况，可以直接ping网站地址
If that can be directly ping a website address to see if there is no echo IP address, as through the barrier is another matter, as long as you can echo the IP address, DNS would be no problem, but there is still this to say about the nslookup command, which It is a very useful command, when I usually own network troubleshooting, basically will be used:
Of course, it can also be used as follows:
Use the nslookup command role is twofold:
1. tests to help you set up your DNS server has no problem
2. Under DNS server irrespective of whether intelligent premise, you can roughly determine the merits of the case according to the speed of DNS server IP addresses echoing speed
So you can take full advantage of nslookup commands.
Fourth, *** IMPORTANT
In fact, basically, a relatively complete network troubleshooting process should be as mentioned above, but, essentially, as explained above, are almost assume that you do not have permission to enter the room to view the case of network equipment.
In fact, you do not need is to copy all of the above steps, the actual troubleshooting process, it may be that you will use just a small part of it, after all, if every so down, no one is not mad, because some very simple questions should be so frustrating, that is a waste of time! However, the above method, but proposes a network troubleshooting thinking, I hope this idea can be judged according to network problem where appearance, as to how to solve, which also involves the middle really too much experience problems (in fact, encountered problems can be Baidu),
In fact, the entire down, I personally feel that the most important is still not the technology itself, but an idea in which the network troubleshooting process, thinking way out decision, which can allow himself to keep a cool head during network troubleshooting, this it's important!