Why can't you be a god?
Many network technology learners, no matter how they listen and practice, but the technology just can't improve, and they can only change careers in the end.
Many network engineering practitioners have 7-8 years of work experience, but when faced with a network construction implementation plan, they are still at a loss, and even get blinded as soon as they leave the simulator.
There are also many peers who have passed the NP and IE level exams, and may not even be equipped with a VLAN ...
So have you thought about it? Is the gap between you and the great god really a CCIE ? I don't think so.
Below, Brother Jie will take stock of a few "bad habits" for everyone, and it is these "bad habits" that will become the biggest stumbling block on your way to becoming a great god... Oh! No, it's a total trap!
One of the vices:
Influenced by books, spend time researching unpopular knowledge
In fact, the Internet is a technology that is easy to "learn and apply", and most of the knowledge points in the book can be used in practical work. But after all, many books are aimed at "meeting exams", so it is inevitable that some unpopular knowledge will appear. Many unpopular knowledge is not only difficult to master, but also has little meaning in actual work. for example:
1. VTP protocol, Cisco proprietary protocol. Although this technology is not difficult to master, in actual work, turning on VTP will bring some difficulties to the management of the local area network, so almost no one uses VTP in actual work .
2. OSPF DR and BDR , unless you encounter the situation of traversing DMVPN to establish OSPF neighbors, otherwise, between the two point-to-point neighbors between the current Layer 3 links, who is the DR and who is the BDR will not affect the routing transfer. So it doesn't really make sense.
3. RIP routing protocol, as a primitive and functional routing protocol, is almost useless in actual work.
So what are the technologies that actually work?
1, OSPF , bear the brunt! You need to master not only the OSPF configuration, but also the six common OSPF LSA query methods (you must read the query commands and query results), and the meaning of the OSPF interface Cost value configuration. There is the redistribution of OSPF and other routes, the concept of OSPF multi-area (must pay attention to the nature of backbone area and non-backbone area).
2. VLAN and Trunk , needless to say, must be able to configure, and must correctly understand the concept of Trunk (many people just can't understand it correctly). When to use Trunk and when to use Access , please think and understand more.
3. Spanning tree ( STP , Spanning-Tree ), you are not required to master any advanced STP technology such as CIST , it is really meaningless. You have a good grasp of the MSTP configuration. When planning the network, the computing area of the spanning tree can be formed into a triangle with three devices.
4. BGP and MPLS VPN , which are the most advanced technologies in routing switching
5. Common link monitoring technologies, such as SLA , NQA , SNMP , etc.
6. ACL , SSH , NAT , etc., these must not only be understood but also configured
7. You must master the content related to route filtering and strategy, such as Prefix -List and Route-map .
The second vice:
Addresses other than 192.168.xx will not be used
To be honest, as soon as Brother Jie saw someone doing an experiment with an address starting with " 192.168 ", he could immediately guess that he was a novice. In fact, there are so many private addresses available, so why develop the "bad habit" of using addresses starting with " 192.168 "?
First of all, I need to explain: The two addresses of 192.168.0.0/24 and 192.168.1.0/24 are called "junk IP ". How can they be "junk"? That is because most brands of wireless routers, the address of the LAN side is 192.168.0.0/24 or 192.168.1.0/24 , so in order to avoid the conflict between the routing in the LAN and the routing of the LAN segment of the wireless router, in the actual network construction process It is strictly forbidden to use 192.168.0.0/24 and 192.168.1.0/24 .
所以，即使是你做小实验，在自行规划IP地址的时候，一定要避开192.168开头的地址，养成坚决不用192.168开头的地址的习惯。
恶习之三：
不分场合的使用子网掩码
随意使用子网掩码，表现出了“小白”在对待实验或者是对待实际环境中的一种不严谨的态度。如果一旦养成习惯，会在走向工作岗位以后，给工作带来不小的麻烦。我就随便举两个例子吧：
比如部分初学者在做实验的时候，设计地址的时候，不管怎样，把/24的地址设置为万能的地址了：
而在实际的网络环境中，尤其是骨干网，两台设备做点对点三层链路连接的时候，一般都会使用/30的子网掩码。所以呢，这个时候如果你在这两端互联接口上使用/24的掩码，必然会造成地址浪费。
虽说都是使用私有地址，但在一些大型网络中，即使是私有地址也是经过了严格的规划的，也不能随便浪费使用。
恶习之四：
随便选一个地址就当了网关
从理论上来讲，一个网段内任何一个地址都可以当网关。比如10.112.10.0/24，从10.112.10.1—254，随便选个地址都能当网关，这都没问题，也都合理合法。但是这违反了网络行业里一个“公序良俗”。捷哥举个很难听的例子，比如你和你老婆离婚，娶了你的小姨子！只要一个愿离一个愿结，这并不违法吧，但这总让人心里觉得膈应得慌。
所以，一个不成文的规定就是：网关地址一般选用一个网段内可用IP地址的最大值或者最小值。例如10.112.10.0/24，担任网关的一般就是10.112.10.1或者10.112.10.254。
恶习之五：
没有配置Loopback地址的习惯
这个，特别是初学者在做实验的时候，觉得Loopback地址配置起来没多大用处，有时候还显得麻烦，但你却不知道，在实际网络环境中，Loopback地址的用途大大的有啊！
比如，做OSPF的时候，用Loopback地址当做Router-id，你可以很清晰的判断出邻居的身份。
加入不用Loopback地址，会造成什么后果呢？
如果此时SW1和SW2形成OSPF完全邻接关系，那么在SW2上，根据Router-id的选举原则，Gi 0/2接口的地址10.41.48.25怎么也大于10.41.48.22吧！于是，SW2的Router-id会被选举为10.41.48.25
如果在SW1上查看OSPF邻居，会发现它的邻居是10.41.48.25！
如果你就看我这个拓扑图，你当然觉得这没什么啊。但如果是一个你不熟悉环境的大型网络呢？Router-id不明确，会加大网络的运维难度的。
而且，Loopback地址除了Router-id以外，还有一个更为重要的作用，那就是应用在BGP/MPLS VPN网络中。
在配置iBGP邻居，或者MPLS的lsr-id时，使用Loopback 地址会比使用物理地址更稳定，我们看图：
如果R1和R2之间，R1使用Gi 0/1的地址和R2的Gi 0/2的地址建立iBGP邻居关系。如果此时R1的Gi 0/1接口down掉或者R2的Gi 0/2接口down掉，则R1和R2的iBGP邻居就直接挂了，没商量！
但如果R1和R2使用Loopback地址建立的iBGP邻居呢？那么即使R1的Gi 0/1接口down掉或者R2的Gi 0/2接口down掉，R1和R2的Loopback 地址还可以通过R1—R3—R2这样子互通啊。
这也就是说Loopback接口的重要性了啊。
所以，在平时做实验的时候，一定要养成配置Loopback接口的习惯。