Improving the efficiency of enterprise routers (1)

The Core routing and switching devices of enterprises shoulder the important responsibilities of connecting clients and servers in the network. Therefore, the operating efficiency of enterprise routers directly determines the operational efficiency of enterprises. At ordinary times, enterprise network administrators always encounter problems such as router Management and Maintenance. Today, I will introduce some effective methods to readers of IT168 network channel from my personal experience, let's use three axes to improve the efficiency of enterprise routers.

I. What is the efficiency of enterprise routers:

Generally, the efficiency of enterprise routers mainly involves the following aspects. The first is the ability of the router itself to process data communication. To be precise, it is the efficiency of data forwarding. Like the server, the router itself has the CPU and memory, when the number of interfaces for communication is large, the CPU and memory load may increase, just like running large programs on a computer, there may be a certain amount of time of stagnation and false death. When a vro encounters a similar situation, network communication may also cause a problem. To improve the running efficiency of the vro, You can minimize the occurrence of the problem or quickly locate the fault when the machine gets stuck frequently. The second step is to manage the configurations of routers. As we all know, the routing switching device has many parameters. In the process of enterprise network operation and maintenance, we need to adjust the parameter configuration according to the actual situation to implement different functions through various commands, how to quickly restore a vro when faulty parameters are lost is also a major factor in determining its operation efficiency, therefore, learning to back up and restore the parameter configuration information is beneficial to improving the efficiency of the routing switch device. The second is the management of the router itself. Generally, there are multiple routing switch devices in the enterprise, how can we manage them effectively? All the operations via telnet are cumbersome, and it is difficult to remember the passwords of each device. Therefore, the overall management of the routing switch device can greatly improve its operation efficiency.

2. Check the load to ensure that the routing switch device runs smoothly:

When we find that the Intranet operation speed of an enterprise slows down and frequent card-mounting problems occur, we should first find the cause from the routing switch device. As far as my experience is concerned, the most effective and simple way to directly view the load of the route exchange device is to identify the CPU load. I will introduce the H3C high-end router as an example, of course, the product operation commands of other vendors are similar, so we will not introduce them here.

Step 1: when the network administrator suspects that the network load is too large, causing the routing switch device to be stuck, he should first watch the various indicators on the routing switch device to confirm that there is no similar broadcast storm. (If a broadcast storm occurs, the CPU load will inevitably exceed 90%, and the indicators on each port will flash wildly)

Step 2: After the link becomes a ring, we use telnet or other management protocols to log on to the management interface of the route switch device.

Step 3: run the super or SYS command to enter the configuration mode, and then run "display cpu". After you press enter, the routing switch device lists the CPU usage of each slot, generally, if the load is lower than 50%, it indicates that the operation is normal. If the load is higher than 70%, we need to identify which slot brings high CPU load, and then find the cause from the connections under the slot. Of course, we can also use the display cpu slot 2 command to view the CPU usage of the specific slot (slot 2. At the same time, the display status lists the CPU usage of the last five seconds, the last one minute, and the last five minutes, helping us understand the time when the problem occurred. (1)

TIPS:

If the network is properly planned, the CPU load is generally not too high, resulting in increased CPU load. In addition to broadcast storms, there may be hardware damage or serious worms on the connected devices.