An example of a very common fault in a hotel network

Environment Introduction:
1. Topology as shown, including wireless network, Poe switch, wired access layer switch "core switch, deep convincing AC, export firewall, etc."
2. This is the hotel's room network, because it is a network, so did not say double redundancy design, normal;
3. Switches are Huawei products, access with 27 57 series, the core with 7706;
4. Each floor has a VLAN design.

Fault Description:
Say things are like, the whole network is I build up, so very clear network environment, the whole network just set up, all floor rooms let people go online testing, yes, everything is normal. Hotel officially opened the second day of trouble appeared: four floor with six floor network appeared very slow situation, customer complaints, first of all we received customer failure problem is the most important to ask clear the following questions:
-What happens when the fault occurs
-A detailed description of the breakdown: How large is the range, part of the machine or large area? How large is the range
-Is there a modified configuration
-whether the fault has regularity, such as the occurrence of certain events during the day, or every other time;
-is the fault concentrated on some switches?
-whether the network environment has changed, such as whether the construction of people have changed the line or something;
-Have you ever had a power-down condition?
-It's important to ask the customer what they know and what they are suggesting.

why ask in detail, because a lot of circumstances you do not ask, the customer will not tell you, with this information to you have a lot of help!



Then I telnet (Customer field Zhuhai) Equipment view
1. First to view the configuration of the floor, configuration no changes;
2. View the performance of the 4, 6F switch, including CPU, memory, temperature and so on, the result is OK;
3. Analysis of the core switch, Access switch log, nothing unusual day---here, log is very helpful
4. View the core switch and the 4, 6F switch interconnect line, display interface XX, port also no abnormal situation

--------------on my way to find fault, another 3 F of the floor network also XX-----------------------
Go on...........
I think this is a strange thing, all network configuration, performance, ports are OK ah, think of here, I think it should not be the network equipment problem, I suspect is the internal and external environment, may be the problem caused by the terminal, and then ...

5. I let the customer own notebook to the floor switch access network, let him test, 4F access network, network very card, ping network basic packet loss 90% have
6. Let the customer see the gateway IP, MAC address, the client sends the IP MAC address of the gateway corresponding to me, I log on the core switch dis int VLAN 4 view VLAN interface IP with Mac, good, a contrast on the situation, the IP address of the corresponding MAC address and the client computer obtained not the same.
7. Modify the VLAN of the computer access switch port, divide to VLAN6, test, the IP address is not the same as the core switch configuration, yes, get the address of the 192.168.0.X network segment, and I configured the 192.168.6.X/24 network segment (Customer requirements)

At this point, the fault has been basically determined----have guests with their own router (more wireless routers), the port has also been connected to the LAN port above.

Fault resolution Method:
First analysis: Because the customer's wireless router LAN port to the Hotel network, The router automatically assigns the 192.168.0.X address, and the gateway is set to the address of the router, so the terminal that gets to this router's assigned address will forward all the data to the router that the customer privately carries, and this kind of home router normally carries 10 users, suddenly one hundred or two hundred people's data comes up, the direct card dies.

troubleshooting Methods The simplest and most economical is the "DHCP snooping" basic mainstream network products are supported
1. Enable DHCP snooping on the access switch
2. Access switch and core switch interconnect Trunk configuration Trust
3. Other interfaces enable DHCP snooping

DHCP Enable
DHCP snooping enable IPv4//enable global DHCP snooping functionality and configure the device to process only DHCPV4 messages, saving device CPU utilization

Interface Gigabitethernet 0/0/2
DHCP snooping enable
Quit

Interface Gigabitethernet 0/0/52
DHCP Snooping trusted


Then restart all Access layer switches, world peace!

In fact, there is a method, is to let each room customers use PPPoE dialing, so that solves this problem, the City hot equipment (non-advertising) is very convenient to fix, can understand this goods, used before, of course, is the purchase of additional equipment, but the customer bought does not increase profits, haha ha!

Transferred from: http://bbs.51cto.com/thread-1377271-1.html

An example of a very common fault in a hotel network