About DHCP principles

About DHCP principles

DHCP principle IP address lease request packets sent by the customer
The DHCP Client initializes TCP/IP and sends a dhcpdiscover broadcast packet to the network through UDP port 67 to request the lease IP address. The Source IP address in the broadcast package is 0.0.0.0, and the target IP address is 255.255.255.255. The package also contains the client's MAC address and computer name.
IP address lease in DHCP response provides packets
Any DHCP server that receives a dhcpdiscover broadcast packet and can provide an IP address will respond to a dhcpoffer broadcast packet to the client through UDP port 68, providing an IP address. The source IP address of the broadcast package is the DCHP Server IP address, and the target IP address is 255.255.255.255. The package also contains the provided IP address, subnet mask, and lease period.
Customer selects IP lease packets
After receiving the provision from more than one DHCP server, the client selects the first received dhcpoffer package and broadcasts a dhcprequest message package to the network, indicates that you have accepted an IP address provided by the DHCP server. The broadcast package contains the accepted IP address and the IP address of the server. All other DHCP servers cancel their provision so that the IP address can be provided to the next IP address lease request.
The DHCP server sends an IP lease confirmation message.
After receiving the dhcprequest broadcast, the DHCP server selected by the client returns a dhcpack message packet to the client, indicating that the client has been selected, add the legal lease and other configuration information of this IP address to the broadcast package and send it to the client.
Announcement message sent after successful configuration by the customer
When the client receives the dhcpack package, it uses the information in the broadcast package to configure its own TCP/IP, the lease process is complete, and the client can communicate in the network.
At this point, the DHCP service process for a customer to obtain an IP address is basically over. However, the IP address obtained by the customer generally uses a lease period, and the lease period needs to be updated before expiration. This process is completed by renting Update packets.
Customer IP address lease update message

We know that in TCP/IP network applications, the network user PC can communicate with other network users only after obtaining a network address. In actual applications, we often encounter some problems: for example, if an IP address conflict occurs, the gateway or DNS server address is incorrectly set, and other hosts in the network cannot be accessed, the IP address must be modified frequently due to frequent location changes. Based on these problems in network management, the solution is to introduce the DHCP service to dynamically configure the information of the customer's machines. The following is based on the DHCP principle and my practical experience in the teaching process, in-depth discussion of DHCP Application and various problems and corresponding solutions in practical application.

1. What is the DHCP service? DHCP is the Dynamic Host Configuration Protocol. The Dhcp Service allows the workstation to connect to the network and automatically obtain an IP address. The server configured with the DHCP service can provide each network customer with an IP address, subnet mask, default gateway, IP address of a WINS server, and IP address of a DNS server.

Ii. FAQs about DHCP service in practical application

1. Can Multiple DHCP servers exist in a subnetwork? If so, can the clients in this subnet correctly obtain the IP address and obtain the IP address assigned by the DHCP server, can the customer machine be controlled to obtain addresses from DHCP servers set by the management personnel rather than obtaining illegal IP addresses from some self-built DHCP servers?

2. If multiple subnets exist in the network, and subnet clients require the DHCP server to provide address configuration, a DHCP server is installed on each subnet, or how should I install a DHCP server in a subnet to assign IP addresses to multiple subnet clients?

3. If a DHCP server is installed in a subnet to assign IP addresses to multiple subnet clients, you must create multiple scopes in different ranges on a DHCP server, but how can we ensure that the IP addresses in the corresponding range are allocated to the corresponding subnet host?

4. If the client machine cannot obtain the IP address from the DHCP server, how will the Windows2000 client machine handle its TCP/IP settings?

Iii. How DHCP works

To resolve the problems raised in the second point, first of all, we need to understand the actual working process and principles of DHCP. Next we will briefly introduce this: DHCP is a broadcast-based protocol, its operations can be divided into four stages: IP leasing requests, IP leasing provision, IP leasing selection, and IP leasing confirmation.

1. IP address lease request: if the customer's computer is set to automatically obtain the IP address at any time, it will check whether it has rented an IP address when it starts up. If not, it requests a lease from DCHP. Because the client computer does not know the address of the DHCP server, 255.255.255.255 is used as the destination address, and 0.0.0.0 is used as the source address to broadcast a dhcpdiscover message on the network, the message contains the Media Access Control (MAC) address of the client computer (the hardware address built in the NIC) and its NetBIOS name.

2. IP Leasing: When the DHCP server receives an IP leasing request from a customer, it will reserve an IP address for the customer based on its own scope address pool and broadcast one on the network, the message contains the customer's MAC address, the IP address that the server can provide, the subnet mask, the lease term, and the IP address of the rented DHCP server.

3. IP Leasing: if another DHCP server exists in the subnet, after the client receives the dhcpoffer message from a DHCP server, it broadcasts a dhcprequest message containing the IP address of the server that provides the lease. In this subnet, it notifies all other DHCP servers that it has accepted the provision of an IP address, after receiving this message, other DHCP servers will unlease the client. Then, return the lease address assigned to the customer to the address pool. The address can be used as a valid address for other computers.

4. IP lease confirmation: When the DHCP server receives the dhcprequest message from the customer, it starts the last phase of the configuration process. The DHCP server sends a dhcpack packet to the customer, this package includes a lease term and all other configuration information requested by the customer. So far, the TCP/IP configuration has been completed.

Iv. solutions to common problems of the DHCP service

In the second part of this article, we have raised four points. Below we will analyze the four problems and provide some personal solutions, it is believed that it will be helpful to the students who learn the DHCP service. These questions are also the most difficult and Difficult Points for students in practical teaching.

1. According to the IP address leasing principle of the customer's computer, if Multiple DHCP servers provide IP Address Configuration Information in a sub-network, this does not violate the request and allocation principles, as long as one client computer initiates an IP address lease request in this subnet, as the request is broadcast, there can be any number of specific DHCP servers in the subnet to respond to an IP address lease request, the customer selects one of these specific DHCP servers after the request. This option is random, but note that the customer can only accept one lease for each network interface card, the above-mentioned specific DHCP servers refer to system-authorized DHCP servers, but unauthorized DHCP servers cannot provide normal Address Allocation services in the network. This is very important, in this way, the customer's machine will only obtain the address in the address range set by the Administrator. This new feature of the Windows 2000 operating system, the DHCP server authentication prevents IP address conflicts caused by illegal IP Address allocation by the DHCP server. A major cause of network paralysis and failure to communicate normally. Based on the above analysis, we can conclude that Multiple DHCP servers can be used in a sub-network to provide address allocation, but not any DHCP server can be used as the address, it should be those that have been authenticated by the system. When the client requests the address, it is randomly obtained from the authenticated machine. In addition, from another perspective, using Multiple DHCP servers on the same subnet will provide more fault tolerance capabilities for the DHCP clients it serves, as shown in, in a subnet (192.168.1.0 mask is 255.255.255.0), two DHCP servers are shared. If DHCP server 1 is unavailable, DHCP server 2 can replace it and continue to lease a new address or renew an existing client. The recommended solution is to use the 80/20 rule to divide the scope addresses between two DHCP servers. The specific method is to configure Server 1 to use most addresses (about 80% ), server 2 can be configured to allow clients to use other addresses (about 20% ).

2. If multiple subnets exist in a network, and hosts with multiple subnets need DHCP servers to provide address configuration information, in this case, we can install a DHCP server in each subnet to assign IP addresses to each subnet. However, we generally do not do this in order to save resources, you can install the DHCP server in a subnet so that it can assign IP addresses to multiple subnets. Multi-subnet Address allocation can be achieved through the DHCP relay proxy function, the device used as a relay proxy can be a Windows2000 Server that provides a relay proxy or a router that complies with rfc1542, DHCP/Bootp relay agent (DHCP relay agent can broadcast DHCP/Bootp broadcast information from one CIDR block to another ). The management LAN is divided into three subnets and connected by a Win2000 Server. The following is a specific solution to use DHCP servers across subnets:

(1) install the DHCP relay agent. In the "route and remote access" Window of the Windows2000 Server, expand "Local Server> IP Route Selection> General" in sequence, right-click the "General" option, select "add route protocol" in the pop-up menu, and then select "DHCP relay agent" in the "New route protocol" window ", click "OK.

(2) specify a DHCP server: Right-click the added "DHCP relay agent" option and select "properties" from the menu to go to the "DHCP relay agent properties" dialog box, on the "General" tab, enter the IP address of the DHCP server in subnet 1: 192.168.1.2, click "add", and click "OK" to close the dialog box.

(3) configure the access interface: Right-click the "DHCP relay agent" option and select "add interface" from the menu ", in the "interface" list box in the "New interface of the DHCP relay agent" dialog box, select the interface that can access the DHCP server in subnet 1. The newly added interfaces here should be interface 2 and Interface 3, click "OK. Then, in the displayed "DHCP relay station attributes" dialog box, select the "relay DHCP packet" option, enable its relay function, and click "OK.

(4) configure a super scope on the DHCP server, which contains three common scopes. The scope address ranges can be set to 192.168.1.10 ~ 192.168.1.254 (allocated to the PC of subnet 1); 192.168.2.10 ~ 192.168.2.254 (allocated to the PC of subnet 2); 192.168.3.10 ~ 192.168.3.254 (used by the PC assigned to subnet 3), you must remember that DHCP can only assign a range to each subnet. After the preceding configuration, DHCP clients PC2 and PC3 in subnet 2 and subnet 3 can access the DHCP server in subnet 1 through the DHCP relay agent of host.

3. After a single DHCP server assigns IP addresses to multiple subnets, we need to clarify the problem, if a subnet's PC, such as PC2 in subnet 2 or PC3 in subnet 3, sends an address request, host a can act as a relay proxy to convey the DHCP server in subnet 1 to their requests, but how can the DHCP server determine and accurately allocate the IP address range 192.168.2.0 to PC2 and the IP address range 192.168.3.0 to PC3? This is a question that most students may have. To clarify this question, you can refer to the following principle analysis to find the answer:

Take host PC2 in subnet 2 as an example. DHCP Client PC2 broadcasts the DHCP/Bootp discover message (dhcpdiscover) on subnet 2, and broadcasts the message to UDP (User datamprotocol) packets are sent through port 67. When relay agent host a receives the message, it checks the gateway IP address contained in the message header, if the gateway IP address is 0.0.0.0, replace it with the IP address 192.168.2.1 of interface 2 of relay agent host, then, it is forwarded to subnet 1 where the DHCP server is located (host a also acts as the router function ). When the DHCP server in subnet 1 receives the message, it checks the gateway IP address in the message and determines whether the gateway IP address is included in a scope of DHCP, so as to determine whether it can use the corresponding IP address range to provide IP address lease, of course, in this example, the DHCP server will be from the scope of 192.168.2.10 ~ 192.168.2.254 select an IP address to configure PC2. That is to say, the gateway IP address (giaddr) in the request address message of the DHCP client is the basis for the DHCP server to determine which IP address to select from the DHCP range to configure the client.

4. If the DHCP client cannot find the DHCP server, it selects an IP address from the class B CIDR Block 169.254.0.0 retained by Microsoft as its own IP address. The subnet mask is 255.255.0.0, the IP address selected by the DHCP client uses ARP broadcast to determine whether the selected IP address has been used by other devices on the network. If the IP address has been used, then the customer can select another IP address to perform the test again, and at most 10 IP addresses can be retried until the configuration is obtained successfully. After that, the customer will attempt to communicate with the DHCP server every five minutes in the background. Once the client gets in touch with the server, the client will discard the automatically set IP address, use the IP address allocated by the server and other configuration information. So when you find that the IP address of your PC is a Class B address 169.254.0.0/16 CIDR block one day, you should know what the problem is.

So far, we have analyzed several problems concerning the application of the DHCP service and provided corresponding solutions. We have encountered many other problems in the teaching of the DHCP service, for example, do the IP addresses of the DHCP server have to be statically specified by the Administrator? What are the consequences of setting the IP address to be automatically obtained? Why does the customer request an IP address from the DHCP server called address lease? How long does the lease time help the management and application of IP addresses on the network? How can I use the DHCP service to enable a host that automatically obtains an IP address to obtain a specified fixed address at any time? All these questions require teachers to carefully explore the knowledge points of each detail of the Service in their teaching so that they can thoroughly teach the principles of the knowledge points and their practical applications, it is also a prerequisite for us to master and use the DHCP service.

DHCP principle IP address lease request packets sent by the customer
The DHCP Client initializes TCP/IP and sends a dhcpdiscover broadcast packet to the network through UDP port 67 to request the lease IP address. The Source IP address in the broadcast package is 0.0.0.0, and the target IP address is 255.255.255.255. The package also contains the client's MAC address and computer name.
IP address lease in DHCP response provides packets
Any DHCP server that receives a dhcpdiscover broadcast packet and can provide an IP address will respond to a dhcpoffer broadcast packet to the client through UDP port 68, providing an IP address. The source IP address of the broadcast package is the DCHP Server IP address, and the target IP address is 255.255.255.255. The package also contains the provided IP address, subnet mask, and lease period.
Customer selects IP lease packets
After receiving the provision from more than one DHCP server, the client selects the first received dhcpoffer package and broadcasts a dhcprequest message package to the network, indicates that you have accepted an IP address provided by the DHCP server. The broadcast package contains the accepted IP address and the IP address of the server. All other DHCP servers cancel their provision so that the IP address can be provided to the next IP address lease request.
The DHCP server sends an IP lease confirmation message.
After receiving the dhcprequest broadcast, the DHCP server selected by the client returns a dhcpack message packet to the client, indicating that the client has been selected, add the legal lease and other configuration information of this IP address to the broadcast package and send it to the client.
Announcement message sent after successful configuration by the customer
When the client receives the dhcpack package, it uses the information in the broadcast package to configure its own TCP/IP, the lease process is complete, and the client can communicate in the network.
At this point, the DHCP service process for a customer to obtain an IP address is basically over. However, the IP address obtained by the customer generally uses a lease period, and the lease period needs to be updated before expiration. This process is completed by renting Update packets.
Customer IP address lease update message

We know that in TCP/IP network applications, the network user PC can communicate with other network users only after obtaining a network address. In actual applications, we often encounter some problems: for example, if an IP address conflict occurs, the gateway or DNS server address is incorrectly set, and other hosts in the network cannot be accessed, the IP address must be modified frequently due to frequent location changes. Based on these problems in network management, the solution is to introduce the DHCP service to dynamically configure the information of the customer's machines. The following is based on the DHCP principle and my practical experience in the teaching process, in-depth discussion of DHCP Application and various problems and corresponding solutions in practical application.

1. What is the DHCP service? DHCP is the Dynamic Host Configuration Protocol. The Dhcp Service allows the workstation to connect to the network and automatically obtain an IP address. The server configured with the DHCP service can provide each network customer with an IP address, subnet mask, default gateway, IP address of a WINS server, and IP address of a DNS server.

Ii. FAQs about DHCP service in practical application

1. Can Multiple DHCP servers exist in a subnetwork? If so, can the clients in this subnet correctly obtain the IP address and obtain the IP address assigned by the DHCP server, can the customer machine be controlled to obtain addresses from DHCP servers set by the management personnel rather than obtaining illegal IP addresses from some self-built DHCP servers?

2. If multiple subnets exist in the network, and subnet clients require the DHCP server to provide address configuration, a DHCP server is installed on each subnet, or how should I install a DHCP server in a subnet to assign IP addresses to multiple subnet clients?

3. If a DHCP server is installed in a subnet to assign IP addresses to multiple subnet clients, you must create multiple scopes in different ranges on a DHCP server, but how can we ensure that the IP addresses in the corresponding range are allocated to the corresponding subnet host?

4. If the client machine cannot obtain the IP address from the DHCP server, how will the Windows2000 client machine handle its TCP/IP settings?

Iii. How DHCP works

To resolve the problems raised in the second point, first of all, we need to understand the actual working process and principles of DHCP. Next we will briefly introduce this: DHCP is a broadcast-based protocol, its operations can be divided into four stages: IP leasing requests, IP leasing provision, IP leasing selection, and IP leasing confirmation.

1. IP address lease request: if the customer's computer is set to automatically obtain the IP address at any time, it will check whether it has rented an IP address when it starts up. If not, it requests a lease from DCHP. Because the client computer does not know the address of the DHCP server, 255.255.255.255 is used as the destination address, and 0.0.0.0 is used as the source address to broadcast a dhcpdiscover message on the network, the message contains the Media Access Control (MAC) address of the client computer (the hardware address built in the NIC) and its NetBIOS name.

2. IP Leasing: When the DHCP server receives an IP leasing request from a customer, it will reserve an IP address for the customer based on its own scope address pool and broadcast one on the network, the message contains the customer's MAC address, the IP address that the server can provide, the subnet mask, the lease term, and the IP address of the rented DHCP server.

3. IP Leasing: if another DHCP server exists in the subnet, after the client receives the dhcpoffer message from a DHCP server, it broadcasts a dhcprequest message containing the IP address of the server that provides the lease. In this subnet, it notifies all other DHCP servers that it has accepted the provision of an IP address, after receiving this message, other DHCP servers will unlease the client. Then, return the lease address assigned to the customer to the address pool. The address can be used as a valid address for other computers.

4. IP lease confirmation: When the DHCP server receives the dhcprequest message from the customer, it starts the last phase of the configuration process. The DHCP server sends a dhcpack packet to the customer, this package includes a lease term and all other configuration information requested by the customer. So far, the TCP/IP configuration has been completed.

Iv. solutions to common problems of the DHCP service

In the second part of this article, we have raised four points. Below we will analyze the four problems and provide some personal solutions, it is believed that it will be helpful to the students who learn the DHCP service. These questions are also the most difficult and Difficult Points for students in practical teaching.

1. According to the IP address leasing principle of the customer's computer, if Multiple DHCP servers provide IP Address Configuration Information in a sub-network, this does not violate the request and allocation principles, as long as one client computer initiates an IP address lease request in this subnet, as the request is broadcast, there can be any number of specific DHCP servers in the subnet to respond to an IP address lease request, the customer selects one of these specific DHCP servers after the request. This option is random, but note that the customer can only accept one lease for each network interface card, the above-mentioned specific DHCP servers refer to system-authorized DHCP servers, but unauthorized DHCP servers cannot provide normal Address Allocation services in the network. This is very important, in this way, the customer's machine will only obtain the address in the address range set by the Administrator. This new feature of the Windows 2000 operating system, the DHCP server authentication prevents IP address conflicts caused by illegal IP Address allocation by the DHCP server. A major cause of network paralysis and failure to communicate normally. Based on the above analysis, we can conclude that Multiple DHCP servers can be used in a sub-network to provide address allocation, but not any DHCP server can be used as the address, it should be those that have been authenticated by the system. When the client requests the address, it is randomly obtained from the authenticated machine. In addition, from another perspective, using Multiple DHCP servers on the same subnet will provide more fault tolerance capabilities for the DHCP clients it serves, as shown in, in a subnet (192.168.1.0 mask is 255.255.255.0), two DHCP servers are shared. If DHCP server 1 is unavailable, DHCP server 2 can replace it and continue to lease a new address or renew an existing client. The recommended solution is to use the 80/20 rule to divide the scope addresses between two DHCP servers. The specific method is to configure Server 1 to use most addresses (about 80% ), server 2 can be configured to allow clients to use other addresses (about 20% ).

2. If multiple subnets exist in a network, and hosts with multiple subnets need DHCP servers to provide address configuration information, in this case, we can install a DHCP server in each subnet to assign IP addresses to each subnet. However, we generally do not do this in order to save resources, you can install the DHCP server in a subnet so that it can assign IP addresses to multiple subnets. Multi-subnet Address allocation can be achieved through the DHCP relay proxy function, the device used as a relay proxy can be a Windows2000 Server that provides a relay proxy or a router that complies with rfc1542, DHCP/Bootp relay agent (DHCP relay agent can broadcast DHCP/Bootp broadcast information from one CIDR block to another ). The management LAN is divided into three subnets and connected by a Win2000 Server. The following is a specific solution to use DHCP servers across subnets:

(1) install the DHCP relay agent. In the "route and remote access" Window of the Windows2000 Server, expand "Local Server> IP Route Selection> General" in sequence, right-click the "General" option, select "add route protocol" in the pop-up menu, and then select "DHCP relay agent" in the "New route protocol" window ", click "OK.

(2) specify a DHCP server: Right-click the added "DHCP relay agent" option and select "properties" from the menu to go to the "DHCP relay agent properties" dialog box, on the "General" tab, enter the IP address of the DHCP server in subnet 1: 192.168.1.2, click "add", and click "OK" to close the dialog box.

(3) configure the access interface: Right-click the "DHCP relay agent" option and select "add interface" from the menu ", in the "interface" list box in the "New interface of the DHCP relay agent" dialog box, select the interface that can access the DHCP server in subnet 1. The newly added interfaces here should be interface 2 and Interface 3, click "OK. Then, in the displayed "DHCP relay station attributes" dialog box, select the "relay DHCP packet" option, enable its relay function, and click "OK.

(4) configure a super scope on the DHCP server, which contains three common scopes. The scope address ranges can be set to 192.168.1.10 ~ 192.168.1.254 (allocated to the PC of subnet 1); 192.168.2.10 ~ 192.168.2.254 (allocated to the PC of subnet 2); 192.168.3.10 ~ 192.168.3.254 (used by the PC assigned to subnet 3), you must remember that DHCP can only assign a range to each subnet. After the preceding configuration, DHCP clients PC2 and PC3 in subnet 2 and subnet 3 can access the DHCP server in subnet 1 through the DHCP relay agent of host.

3. After a single DHCP server assigns IP addresses to multiple subnets, we need to clarify the problem, if a subnet's PC, such as PC2 in subnet 2 or PC3 in subnet 3, sends an address request, host a can act as a relay proxy to convey the DHCP server in subnet 1 to their requests, but how can the DHCP server determine and accurately allocate the IP address range 192.168.2.0 to PC2 and the IP address range 192.168.3.0 to PC3? This is a question that most students may have. To clarify this question, you can refer to the following principle analysis to find the answer:

Take host PC2 in subnet 2 as an example. DHCP Client PC2 broadcasts the DHCP/Bootp discover message (dhcpdiscover) on subnet 2, and broadcasts the message to UDP (User datamprotocol) packets are sent through port 67. When relay agent host a receives the message, it checks the gateway IP address contained in the message header, if the gateway IP address is 0.0.0.0, replace it with the IP address 192.168.2.1 of interface 2 of relay agent host, then, it is forwarded to subnet 1 where the DHCP server is located (host a also acts as the router function ). When the DHCP server in subnet 1 receives the message, it checks the gateway IP address in the message and determines whether the gateway IP address is included in a scope of DHCP, so as to determine whether it can use the corresponding IP address range to provide IP address lease, of course, in this example, the DHCP server will be from the scope of 192.168.2.10 ~ 192.168.2.254 select an IP address to configure PC2. That is to say, the gateway IP address (giaddr) in the request address message of the DHCP client is the basis for the DHCP server to determine which IP address to select from the DHCP range to configure the client.

4. If the DHCP client cannot find the DHCP server, it selects an IP address from the class B CIDR Block 169.254.0.0 retained by Microsoft as its own IP address. The subnet mask is 255.255.0.0, the IP address selected by the DHCP client uses ARP broadcast to determine whether the selected IP address has been used by other devices on the network. If the IP address has been used, then the customer can select another IP address to perform the test again, and at most 10 IP addresses can be retried until the configuration is obtained successfully. After that, the customer will attempt to communicate with the DHCP server every five minutes in the background. Once the client gets in touch with the server, the client will discard the automatically set IP address, use the IP address allocated by the server and other configuration information. So when you find that the IP address of your PC is a Class B address 169.254.0.0/16 CIDR block one day, you should know what the problem is.

So far, we have analyzed several problems concerning the application of the DHCP service and provided corresponding solutions. We have encountered many other problems in the teaching of the DHCP service, for example, do the IP addresses of the DHCP server have to be statically specified by the Administrator? What are the consequences of setting the IP address to be automatically obtained? Why does the customer request an IP address from the DHCP server called address lease? How long does the lease time help the management and application of IP addresses on the network? How can I use the DHCP service to enable a host that automatically obtains an IP address to obtain a specified fixed address at any time? All these questions require teachers to carefully explore the knowledge points of each detail of the Service in their teaching so that they can thoroughly teach the principles of the knowledge points and their practical applications, it is also a prerequisite for us to master and use the DHCP service.