IP Security (IPsec) is defined by the IETF to provide interoperable, high-quality, cryptography-based security for IP communications. It is a Layer 3 VPN technology that transmits data in a secure channel established between two endpoints (such as two security gateways). Such a secure channel is usually called an IPsec tunnel.

IPsec framework

IPsec is a security framework that has the following protocols and algorithms:

· Authentication Header (AH).

· Encapsulating Security Payload (ESP).

· Internet Key Exchange (IKE).

· Algorithms for authentication and encryption.

AH and ESP are security protocols that provide security services. IKE performs automatic key exchange. For more information about IKE, see "Configuring IKE."

IPsec security services

IPsec provides the following security services for data packets in the IP layer:

· Confidentiality—The sender encrypts packets before transmitting them over the Internet, protecting the packets from being eavesdropped en route.

· Data integrity—The receiver verifies the packets received from the sender to make sure they are not tampered with during transmission.

· Data origin authentication—The receiver verifies the authenticity of the sender.

· Anti-replay—The receiver examines packets and drops outdated and duplicate packets.

Benefits of IPsec

IPsec delivers the following benefits:

· Reduced key negotiation overhead and simplified maintenance by supporting the IKE protocol. IKE provides automatic key negotiation and automatic IPsec security association (SA) setup and maintenance.

· Good compatibility. You can apply IPsec to all IP-based application systems and services without modifying them.

· Encryption on a per-packet rather than per-flow basis. Per-packet encryption allows for flexibility and greatly enhances IP security.

Security protocols

IPsec comes with two security protocols, AH and ESP. They define how to encapsulate IP packets and the security services that they can provide.

· AH (protocol 51) defines the encapsulation of the AH header in an IP packet, as shown in Figure 3. AH can provide data origin authentication, data integrity, and anti-replay services to prevent data tampering, but it cannot prevent eavesdropping. Therefore, it is suitable for transmitting non-confidential data. Authentication algorithms supported by AH include HMAC-MD5 and HMAC-SHA1.

AH does not support NAT traversal.

· ESP (protocol 50) defines the encapsulation of the ESP header and trailer in an IP packet, as shown in Figure 3. ESP can provide data encryption, data origin authentication, data integrity, and anti-replay services. Unlike AH, ESP can guarantee data confidentiality because it can encrypt the data before encapsulating the data to IP packets. ESP-supported encryption algorithms include DES, 3DES, and AES, and authentication algorithms include HMAC-MD5 and HMAC-SHA1.

Both AH and ESP provide authentication services, but the authentication service provided by AH is stronger. In practice, you can choose either or both security protocols. When both AH and ESP are used, an IP packet is encapsulated first by ESP and then by AH.

Encapsulation modes

IPsec supports the following encapsulation modes: transport mode and tunnel mode.

Transport mode

The security protocols protect the upper layer data of an IP packet. Only the transport layer data is used to calculate the security protocol headers. The calculated security protocol headers and the encrypted data (only for ESP encapsulation) are placed after the original IP header. You can use the transport mode when end-to-end security protection is required (the secured transmission start and end points are the actual start and end points of the data). The transport mode is typically used for protecting host-to-host communications, as shown in Figure 1.

Figure 1 IPsec protection in transport mode

Tunnel mode

The security protocols protect the entire IP packet. The entire IP packet is used to calculate the security protocol headers. The calculated security protocol headers and the encrypted data (only for ESP encapsulation) are encapsulated in a new IP packet. In this mode, the encapsulated packet has two IP headers. The inner IP header is the original IP header. The outer IP header is added by the network device that provides the IPsec service. You must use the tunnel mode when the secured transmission start and end points are not the actual start and end points of the data packets (for example, when two gateways provide IPsec but the data start and end points are two hosts behind the gateways). The tunnel mode is typically used for protecting gateway-to-gateway communications, as shown in Figure 2.

Figure 2 IPsec protection in tunnel mode

Figure 3 shows how the security protocols encapsulate an IP packet in different encapsulation modes.

Figure 3 Security protocol encapsulations in different modes

Security association

About this task

A security association (SA) is an agreement negotiated between two communicating parties called IPsec peers. An SA includes the following parameters for data protection:

· Security protocols (AH, ESP, or both).

· Encapsulation mode (transport mode or tunnel mode).

· Authentication algorithm (HMAC-MD5 or HMAC-SHA1).

· Encryption algorithm (DES, 3DES, or AES).

· Shared keys and their lifetimes.

An SA is unidirectional. At least two SAs are needed to protect data flows in a bidirectional communication. If two peers want to use both AH and ESP to protect data flows between them, they construct an independent SA for each protocol in each direction.

An SA is uniquely identified by a triplet, which consists of the security parameter index (SPI), destination IP address, and security protocol identifier. An SPI is a 32-bit number. It is transmitted in the AH/ESP header.

SA setup

An SA can be set up manually or through IKE.

· Manual mode—Configure all parameters for the SA through commands. This configuration mode is complex and does not support some advanced features (such as periodic key update), but it can implement IPsec without IKE. This mode is mainly used in small and static networks or when the number of IPsec peers in the network is small.

· IKE negotiation mode—The peers negotiate and maintain the SA through IKE. This configuration mode is simple and has good expansibility. As a best practice, set up SAs through IKE negotiations in medium- and large-scale dynamic networks.

SA aging

A manually configured SA never ages out.

An IKE-created SA has a lifetime and will be deleted when its lifetime timer expires.

Before the SA lifetime timer expires, IKE negotiates a new SA, which takes over immediately after its creation. The interval from the creation of an SA to the negotiation of a new SA is the SA's soft lifetime.

The SA soft lifetime is calculated as follows: SA soft lifetime = SA lifetime – SA soft lifetime buffer. If the SA soft lifetime buffer is not configured, the system calculates a default SA soft lifetime based on the SA lifetime.

The lifetime of an IKE-created SA comes in two types:

· Time-based lifetime—Defines how long the SA can exist after it is created.

· Traffic-based lifetime—Defines the maximum traffic that the SA can process.

If both lifetime timers are configured for an SA, the SA is deleted when either of the lifetime timers expires.

Authentication and encryption

Authentication algorithms

IPsec uses hash algorithms to perform authentication. A hash algorithm produces a fixed-length digest for an arbitrary-length message. IPsec peers respectively calculate message digests for each packet. The receiver compares the local digest with that received from the sender. If the digests are identical, the receiver considers the packet intact and the sender's identity valid. IPsec supports the following types of authentication algorithms:

· Hash-based Message Authentication Code (HMAC) based authentication algorithms, including HMAC-MD5 and HMAC-SHA.

HMAC-MD5 is faster but less secure than HMAC-SHA.

Encryption algorithms

IPsec uses symmetric encryption algorithms, which encrypt and decrypt data by using the same keys. The following encryption algorithms are available for IPsec on the device:

· DES—Encrypts a 64-bit plaintext block with a 56-bit key. DES is the least secure but the fastest algorithm.

· 3DES—Encrypts plaintext data with three 56-bit DES keys. The key length totals up to 168 bits. It provides moderate security strength and is slower than DES.

· AES—Encrypts plaintext data with a 128-bit, 192-bit, or 256-bit key. AES provides the highest security strength and is slower than 3DES.

Crypto engine

The IPsec feature is resource intensive for its complex encryption/decryption and authentication algorithms. To improve processing performance, you can use crypto engine to offload IPsec tasks.

The crypto engine processes all IPsec protected packets and hands the processed packets back to the device for forwarding.

For more information about crypto engines, see "Configuring crypto engines."

IPsec-protected traffic

IPsec tunnels can protect the following types of traffic:

· Packets that match specific ACLs.

· Packets routed to a tunnel interface.

· Packets of IPv6 routing protocols.

Two peers use IPsec policies to protect packets between them. An IPsec policy defines the range of packets to be protected by IPsec and the security parameters used for the protection. For more information about IPsec policies, see "IPsec policy."

The following information describes how IPsec protects packets:

· When an IPsec peer identifies the packets to be protected according to the security policy, it sets up an IPsec tunnel and sends the packet to the remote peer through the tunnel. The IPsec tunnel can be manually configured beforehand, or it can be set up through IKE negotiation triggered by the packet. The IPsec tunnels are actually the IPsec SAs. The inbound packets are protected by the inbound SA, and the outbound packets are protected by the outbound SA.

· When the remote IPsec peer receives the packet, it drops, de-encapsulates, or directly forwards the packet according to the configured security policy.

ACL-based IPsec

To implement ACL-based IPsec, configure an ACL to define the data flows to be protected, specify the ACL in an IPsec policy, and then apply the IPsec policy to an interface. You can apply an IPsec policy to physical interfaces such as serial interfaces and Ethernet interfaces, or virtual interfaces such as tunnel interfaces and virtual template interfaces.

ACL-based IPsec works as follows:

· When packets sent by the interface match a permit rule of the ACL, the packets are protected by the outbound IPsec SA and encapsulated with IPsec.

· When the interface receives an IPsec packet destined for the local device, it searches for the inbound IPsec SA according to the SPI in the IPsec packet header for de-encapsulation. If the de-encapsulated packet matches a permit rule of the ACL, the device processes the packet. If the de-encapsulated packet does not match any permit rule of the ACL, the device drops the packet.

The device supports the following data flow protection modes:

· Standard mode—One IPsec tunnel protects one data flow. The data flow permitted by an ACL rule is protected by one IPsec tunnel that is established solely for it.

· Aggregation mode—One IPsec tunnel protects all data flows permitted by all the rules of an ACL. This mode is only used to communicate with old-version devices.

· Per-host mode—One IPsec tunnel protects one host-to-host data flow. One host-to-host data flow is identified by one ACL rule and protected by one IPsec tunnel established solely for it. This mode consumes more system resources when multiple data flows exist between two subnets to be protected.

IPsec policy

IPsec policies define the parameters used to establish IPsec tunnels between two peers and the range of packets to be protected.

An IPsec policy is a set of IPsec policy entries that have the same name but different sequence numbers.

An IPsec policy contains the following settings:

· An ACL that defines the range of data flows to be protected.

· An IPsec transform set that defines the security parameters used for IPsec protection.

· IPsec SA establishment mode.

Supported IPsec SA establishment modes are manual configuration and IKE negotiation.

· Local and remote IP addresses that define the start and end points of the IPsec tunnel.

In the same IPsec policy, an IPsec policy entry with a smaller sequence number has a higher priority. When sending a packet, the interface applied with an IPsec policy looks through the IPsec policy's entries in ascending order of sequence numbers. If the packet matches the ACL of an IPsec policy entry, the interface encapsulates the packet according to the IPsec policy entry. If no match is found, the interface sends the packet out without IPsec protection.

When the interface receives an IPsec packet destined for the local device, it searches for the inbound IPsec SA according to the SPI in the IPsec packet header for de-encapsulation. If the de-encapsulated packet matches a permit rule of the ACL, the device processes the packet. If the de-encapsulated packet does not match a permit rule of the ACL, the device drops the packet.

IPsec RRI

IPsec Reverse Route Injection (RRI) enables an IPsec tunnel gateway to automatically add and delete static routes destined for the protected private networks. It automatically adds the static routes when the IPsec SAs are established and deletes the static routes when the IPsec SAs are deleted. This greatly reduces the static route configuration work load on the gateway and increases the scalability of the IPsec VPN.

IPsec RRI is applicable to gateways that must provide many IPsec tunnels (for example, a headquarters gateway).

As shown in Figure 4, the traffic between the enterprise center and the branches are protected by IPsec. The gateway at the enterprise center is configured with static routes to route traffic to the IPsec-protected interfaces. It is difficult to add or modify static routes on the gateway at the enterprise center if the IPsec VPN has a large number of branches or if the network structure changes.

Figure 4 IPsec VPN

After you can enable IPsec RRI on the gateway, the gateway automatically adds a static route to the routing table each time an IPsec tunnel is established. The destination IP address is the protected private network. The next hop IP address can be the remote IP address of the IPsec tunnel (default) or a user-defined next hop IP address. Traffic destined for the peer end is routed to the IPsec tunnel interface and thereby protected by IPsec.

You can advertise the static routes created by IPsec RRI in the internal network, and the internal network device can use them to forward traffic in the IPsec VPN.

You can set preferences for the static routes created by IPsec RRI to implement flexible route management. For example, you can set the same preference for multiple routes to the same destination to implement load sharing, or you can set different preferences to implement route backup.

You can also set tags for the static routes created by IPsec RRI to implement flexible route control through routing policies.

Protocols and standards

· RFC 2401, Security Architecture for the Internet Protocol

· RFC 2402, IP Authentication Header

· RFC 2406, IP Encapsulating Security Payload

· RFC 4552, Authentication/Confidentiality for OSPFv3

Restrictions and guidelines: IPsec configuration

Typically, IKE uses UDP port 500 for communication, and AH and ESP use the protocol numbers 51 and 50, respectively. Make sure traffic of these protocols is not denied on the interfaces with IKE or IPsec configured.

Implementing ACL-based IPsec

ACLs for IPsec take effect only on traffic that is generated by the device and traffic that is destined for the device. They do not take effect on traffic forwarded through the device. For example, an ACL-based IPsec tunnel can protect log messages the device sends to a log server, but it does not protect data flows and voice flows that are forwarded by the device.

ACL-based IPsec tasks at a glance

To configure ACL-based IPsec, perform the following tasks:

1. Configuring an ACL

2. Configuring an IPsec transform set

3. Configuring an IPsec policy

Choose one of the following tasks:

¡ Configuring a manual IPsec policy

¡ Configuring an IKE-based IPsec policy

4. Applying an IPsec policy to an interface

5. (Optional.) Configuring accessibility features for ACL-based IPsec

¡ Enabling ACL checking for de-encapsulated packets

¡ Configuring IPsec anti-replay

¡ Binding a source interface to an IPsec policy

¡ Enabling QoS pre-classify

¡ Configuring the DF bit of IPsec packets

¡ Configuring IPsec RRI

¡ Configuring the global IPsec SA lifetime and idle timeout

¡ Configuring IPsec fragmentation

¡ Setting the maximum number of IPsec tunnels

6. (Optional.) Configuring logging and SNMP notification for IPsec

¡ Enabling logging for IPsec packets

¡ Enabling logging for IPsec negotiation

¡ Configuring SNMP notifications for IPsec

Configuring an ACL

IPsec uses ACLs to identify the traffic to be protected.

Keywords in ACL rules

An ACL is a collection of ACL rules. Each ACL rule is a deny or permit statement. A permit statement identifies a data flow protected by IPsec, and a deny statement identifies a data flow that is not protected by IPsec. IPsec compares a packet against the ACL rules and processes the packet according to the first rule it matches.

· Each ACL rule matches both the outbound traffic and the returned inbound traffic. Suppose there is a rule rule 0 permit ip source 1.1.1.0 0.0.0.255 destination 2.2.2.0 0.0.0.255. This rule matches both traffic from 1.1.1.0 to 2.2.2.0 and traffic from 2.2.2.0 to 1.1.1.0.

· In the outbound direction, if a permit statement is matched, IPsec considers that the packet requires protection and continues to process it. If a deny statement is matched or no match is found, IPsec considers that the packet does not require protection and delivers it to the next module.

· In the inbound direction:

¡ Non-IPsec packets that match a permit statement are dropped.

¡ IPsec packets destined for the device itself are de-encapsulated. By default, the de-encapsulated packets are compared against the ACL rules. Only those that match a permit statement are processed. Other packets are dropped. If ACL checking for de-encapsulated IPsec packets is disabled, the de-encapsulated packets are not compared against the ACL rules and are directly processed by other modules.

When defining ACL rules for IPsec, follow these guidelines:

· Permit only data flows that need to be protected and use the any keyword with caution. With the any keyword specified in a permit statement, all outbound traffic matching the permit statement will be protected by IPsec. All inbound IPsec packets matching the permit statement will be received and processed, but all inbound non-IPsec packets will be dropped. This will cause all the inbound traffic that does not need IPsec protection to be dropped.

· Avoid statement conflicts in the scope of IPsec policy entries. When creating a deny statement, be careful with its match scope and match order relative to permit statements. The policy entries in an IPsec policy have different match priorities. ACL rule conflicts between them are prone to cause mistreatment of packets. For example, when configuring a permit statement for an IPsec policy entry to protect an outbound traffic flow, you must avoid the situation that the traffic flow matches a deny statement in a higher priority IPsec policy entry. Otherwise, the packets will be sent out as normal packets. If they match a permit statement at the receiving end, they will be dropped by IPsec.

The following example shows how an improper statement causes unexpected packet dropping. Only the ACL-related configuration is presented.

Assume Device A is connected to subnet 1.1.2.0/24 and Device B is connected to subnet 3.3.3.0/24, and the IPsec policy configuration on Device A and Device B is as follows:

· IPsec configuration on Device A:

acl advanced 3000

rule 0 permit ip source 1.1.1.0 0.0.0.255 destination 2.2.2.0 0.0.0.255

rule 1 deny ip

acl advanced 3001

rule 0 permit ip source 1.1.2.0 0.0.0.255 destination 3.3.3.0 0.0.0.255

rule 1 deny ip

ipsec policy testa 1 isakmp <---IPsec policy entry with a higher priority

security acl 3000

ike-profile aa

transform-set 1

ipsec policy testa 2 isakmp <---IPsec policy entry with a lower priority

security acl 3001

ike-profile bb

transform-set 1

· IPsec configuration on Device B:

acl advanced 3001

rule 0 permit ip source 3.3.3.0 0.0.0.255 destination 1.1.2.0 0.0.0.255

rule 1 deny ip

ipsec policy testb 1 isakmp

security acl 3001

ike-profile aa

transform-set 1

On Device A, apply the IPsec policy testa to the outbound interface of Device A. The IPsec policy contains two policy entries, testa 1 and testa 2. The ACLs used by the two policy entries each contain a rule that matches traffic from 1.1.2.0/24 to 3.3.3.0/24. The one used in the policy entry testa 1 is a deny statement and the one used in the policy entry testa 2 is a permit statement. Because testa 1 is matched prior to testa 2, traffic from 1.1.2.0/24 to 3.3.3.0/24 will match the deny statement and be sent as normal traffic. When the traffic arrives at Device B, the traffic matches rule 0 (a permit statement) in ACL 3001 used in the applied IPsec policy testb. Because non-IPsec traffic that matches a permit statement must be dropped on the inbound interface, Device B drops the traffic.

To make sure subnet 1.1.2.0/24 can access subnet 3.3.3.0/24, you can delete the deny rule in ACL 3000 on Device A.

Mirror image ACLs

To make sure SAs can be set up and the traffic protected by IPsec can be processed correctly between two IPsec peers, create mirror image ACLs on the IPsec peers. As shown in Figure 5, ACL rules on Device B are mirror images of the rules on Device A. In this way, SAs can be created successfully for the traffic between Host A and Host C and for the traffic between Network 1 and Network 2.

Figure 5 Mirror image ACLs

‌

If the ACL rules on IPsec peers do not form mirror images of each other, SAs can be set up only when both of the following requirements are met:

· The range specified by an ACL rule on one peer is covered by its counterpart ACL rule on the other peer. As shown in Figure 6, the range specified by the ACL rule configured on Device A is covered by its counterpart on Device B.

· The peer with the narrower rule initiates SA negotiation. If a wider ACL rule is used by the SA initiator, the negotiation request might be rejected because the matching traffic is beyond the scope of the responder. As shown in Figure 6, the SA negotiation initiated by Host A to Host C is accepted but the SA negotiations from Host C to Host A, from Host C to Host B, and from Host D to Host A are rejected.

Figure 6 Non-mirror image ACLs

‌

Configuring an IPsec transform set

About this task

An IPsec transform set, part of an IPsec policy, defines the security parameters for IPsec SA negotiation, including the security protocol, encryption algorithms, and authentication algorithms.

Restrictions and guidelines

Changes to an IPsec transform set affect only SAs negotiated after the changes. To apply the changes to existing SAs, execute the reset ipsec sa command to clear the SAs so that they can be set up by using the updated parameters.

The transport mode applies only when the source and destination IP addresses of data flows match those of the IPsec tunnel.

When you configure the Perfect Forward Secrecy (PFS) feature in an IPsec transform set, follow these guidelines:

· In IKEv1, the security level of the DH group of the initiator must be higher than or equal to that of the responder. This restriction does not apply to IKEv2.

· The end without the PFS feature performs SA negotiation according to the PFS requirements of the peer end.

You can specify multiple authentication or encryption algorithms for the same security protocol. The algorithm specified earlier has a higher priority.

Some algorithms are available only for IKEv2. See Table 1.

Table 1 Algorithms available only for IKEv2

Type	Algorithms
Encryption algorithm	aes-ctr-128 aes-ctr-192 aes-ctr-256 camellia-cbc-128 camellia-cbc-192 camellia-cbc-256 gmac-128 gmac-192 gmac-256 gcm-128 gcm-192 gcm-256
Authentication algorithm	aes-xcbc-mac
PFS algorithm	dh-group19 dh-group20

Procedure

1. Enter system view.

system-view

2. Create an IPsec transform set and enter its view.

ipsec transform-set transform-set-name

3. Specify the security protocol for the IPsec transform set.

protocol { ah | ah-esp | esp }

By default, the ESP security protocol is used.

4. Specify the encryption algorithms for ESP. Skip this step if the protocol ah command is configured.

esp encryption-algorithm { 3des-cbc | aes-cbc-128 | aes-cbc-192 | aes-cbc-256 | aes-ctr-128 | aes-ctr-192 | aes-ctr-256 | camellia-cbc-128 | camellia-cbc-192 | camellia-cbc-256 | des-cbc | gmac-128 | gmac-192 | gmac-256 | gcm-128 | gcm-192 | gcm-256 | null } *

By default, no encryption algorithm is specified for ESP.

5. Specify the authentication algorithms for ESP. Skip this step if the protocol ah command is configured.

esp authentication-algorithm { aes-xcbc-mac | md5 | sha1 | sha256 | sha384 | sha512 } *

By default, no authentication algorithm is specified for ESP.

The aes-xcbc-mac algorithm is available only for IKEv2.

6. Specify the authentication algorithms for AH. Skip this step if the protocol esp command is configured.

ah authentication-algorithm { aes-xcbc-mac | md5 | sha1 | sha256 | sha384 | sha512 } *

By default, no authentication algorithm is specified for AH.

The aes-xcbc-mac algorithm is available only for IKEv2.

7. Specify the packet encapsulation mode.

encapsulation-mode { transport | tunnel }

By default, the security protocol encapsulates IP packets in tunnel mode.

8. (Optional.) Enable the PFS feature.

pfs { dh-group1 | dh-group2 | dh-group5 | dh-group14 | dh-group24 | dh-group19 | dh-group20 }

By default, the PFS feature is disabled.

For more information about PFS, see "Configuring IKE."

9. (Optional.) Enable the Extended Sequence Number (ESN) feature.

esn enable [ both ]

By default, the ESN feature is disabled.

Configuring a manual IPsec policy

In a manual IPsec policy, the parameters are configured manually, such as the keys, the SPIs, and the IP addresses of the two ends in tunnel mode.

Restrictions and guidelines

When you configure a manual IPsec policy, make sure the IPsec configuration at both ends of the IPsec tunnel meets the following requirements:

· The IPsec policies at the two ends must have IPsec transform sets that use the same security protocols, security algorithms, and encapsulation mode.

· The remote IPv4 address configured on the local end must be the same as the primary IPv4 address of the interface applied with the IPsec policy at the remote end. The remote IPv6 address configured on the local end must be the same as the first IPv6 address of the interface applied with the IPsec policy at the remote end.

· At each end, configure parameters for both the inbound SA and the outbound SA, and make sure the SAs in each direction are unique: For an outbound SA, make sure its triplet (remote IP address, security protocol, and SPI) is unique. For an inbound SA, make sure its SPI is unique.

· The local inbound SA must use the same SPI and keys as the remote outbound SA. The same is true of the local outbound SA and remote inbound SA.

· The keys for the IPsec SAs at the two tunnel ends must be configured in the same format. For example, if the local end uses a key in hexadecimal format, the remote end must also use a key in hexadecimal format. If you configure a key in both the character and the hexadecimal formats, only the most recent configuration takes effect.

· If you configure a key in character format for ESP, the device automatically generates an authentication key and an encryption key for ESP.

Procedure

1. Enter system view.

system-view

2. Create a manual IPsec policy entry and enter its view.

ipsec { ipv6-policy | policy } policy-name seq-number manual

3. (Optional.) Configure a description for the IPsec policy.

description text

By default, no description is configured.

4. Specify an ACL for the IPsec policy.

security acl [ ipv6 ] { acl-number | name acl-name }

By default, no ACL is specified for an IPsec policy.

You can specify only one ACL for an IPsec policy.

5. Specify an IPsec transform set for the IPsec policy.

transform-set transform-set-name

By default, no IPsec transform set is specified for an IPsec policy.

You can specify only one IPsec transform set for a manual IPsec policy.

6. Specify the remote IP address of the IPsec tunnel.

remote-address { ipv4-address | ipv6 ipv6-address }

By default, the remote IP address of the IPsec tunnel is not specified.

7. Configure an SPI for the inbound IPsec SA.

sa spi inbound { ah | esp } spi-number

By default, no SPI is configured for the inbound IPsec SA.

8. Configure an SPI for the outbound IPsec SA.

sa spi outbound { ah | esp } spi-number

By default, no SPI is configured for the outbound IPsec SA.

9. Configure keys for the IPsec SA.

¡ Configure an authentication key in hexadecimal format for AH.

sa hex-key authentication { inbound | outbound } ah { cipher | simple } string

¡ Configure an authentication key in character format for AH.

sa string-key { inbound | outbound } ah { cipher | simple } string

¡ Configure a key in character format for ESP.

sa string-key { inbound | outbound } esp { cipher | simple } string

¡ Configure an authentication key in hexadecimal format for ESP.

sa hex-key authentication { inbound | outbound } esp { cipher | simple }

¡ Configure an encryption key in hexadecimal format for ESP.

sa hex-key encryption { inbound | outbound } esp { cipher | simple } string

By default, no keys are configured for the IPsec SA.

Configure keys correctly for the security protocol (AH, ESP, or both) you have specified in the IPsec transform set used by the IPsec policy.

Configuring an IKE-based IPsec policy

About this task

In an IKE-based IPsec policy, the parameters are automatically negotiated through IKE.

To configure an IKE-based IPsec policy, use one of the following methods:

· Directly configure it by configuring the parameters in IPsec policy view.

· Configure it by using an existing IPsec policy template with the parameters to be negotiated configured.

A device using an IPsec policy that is configured in this way cannot initiate an SA negotiation, but it can respond to a negotiation request. The parameters not defined in the template are determined by the initiator. For example, in an IPsec policy template, the ACL is optional. If you do not specify an ACL, the IPsec protection range has no limit. So the device accepts all ACL settings of the negotiation initiator.

When the remote end's information (such as the IP address) is unknown, this method allows the remote end to initiate negotiations with the local end.

The configurable parameters for an IPsec policy template are the same as those when you directly configure an IKE-based IPsec policy. The difference is that more parameters are optional for an IPsec policy template. Except the IPsec transform sets and the IKE profile, all other parameters are optional.

Restrictions and guidelines for IKE-based IPsec policy configuration

The IPsec policies at the two tunnel ends must have IPsec transform sets that use the same security protocols, security algorithms, and encapsulation mode.

The IPsec policies at the two tunnel ends must have the same IKE profile parameters.

An IKE-based IPsec policy can use a maximum of six IPsec transform sets. During an IKE negotiation, IKE searches for a fully matched IPsec transform set at the two ends of the IPsec tunnel. If no match is found, no SA can be set up, and the packets expecting to be protected will be dropped.

The remote IP address of the IPsec tunnel is required on an IKE negotiation initiator and is optional on the responder. The remote IP address specified on the local end must be the same as the local IP address specified on the remote end.

The IPsec SA uses the local lifetime settings or those proposed by the peer, whichever are smaller.

The IPsec SA can have both a time-based lifetime and a traffic-based lifetime. The IPsec SA expires when either lifetime expires.

If you specify both an IKEv1 profile and an IKEv2 profile for an IPsec policy, the IKEv2 profile is used preferentially. For more information about IKEv1 and IKEv2 profiles, see "Configuring IKE" and "Configuring IKEv2."

Directly configuring an IKE-based IPsec policy

1. Enter system view.

system-view

2. Create an IKE-based IPsec policy entry and enter its view.

ipsec { ipv6-policy | policy } policy-name seq-number isakmp

3. (Optional.) Configure a description for the IPsec policy.

description text

By default, no description is configured.

4. (Optional.) Set the IPsec SA negotiation triggering mode.

sa trigger-mode { auto | traffic-based }

By default, IPsec SA negotiation is triggered when traffic requires IPsec protection.

5. Specify an ACL for the IPsec policy.

security acl [ ipv6 ] { acl-number | name acl-name } [ aggregation | per-host ]

By default, no ACL is specified for an IPsec policy.

You can specify only one ACL for an IPsec policy.

6. Specify IPsec transform sets for the IPsec policy.

transform-set transform-set-name&<1-6>

By default, no IPsec transform sets are specified for an IPsec policy.

7. Specify an IKE profile or IKEv2 profile for the IPsec policy.

¡ Specify an IKE profile.

ike-profile profile-name

By default, no IKE profile is specified for an IPsec policy.

¡ Specify an IKEv2 profile.

ikev2-profile profile-name

By default, no IKEv2 profile is specified for an IPsec policy.

8. Specify the local IP address of the IPsec tunnel.

local-address { ipv4-address | ipv6 ipv6-address }

By default, the local IPv4 address of the IPsec tunnel is the primary IPv4 address of the interface to which the IPsec policy is applied. The local IPv6 address of the IPsec tunnel is the first IPv6 address of the interface to which the IPsec policy is applied.

The local IP address specified by this command must be the same as the IP address used as the local IKE identity.

In a VRRP network, the local IP address must be the virtual IP address of the VRRP group to which the IPsec-applied interface belongs.

The local address cannot be a secondary IP address of the interface where the IPsec policy is applied.

9. Specify the remote IP address of the IPsec tunnel.

remote-address { [ ipv6 ] host-name | ipv4-address | ipv6 ipv6-address }

By default, the remote IP address of the IPsec tunnel is not specified.

10. (Optional.) Set the lifetime, soft lifetime buffer, or idle timeout for the IPsec SA.

¡ Set the IPsec SA lifetime.

sa duration { time-based seconds | traffic-based kilobytes }

By default, the global SA lifetime is used.

¡ Set the time-based or traffic-based IPsec SA soft lifetime buffer.

sa soft-duration buffer { time-based seconds | traffic-based kilobytes }

By default, no IPsec SA soft lifetime buffers are configured.

¡ Set the IPsec SA idle timeout.

sa idle-time seconds

By default, the global IPsec SA idle timeout is used.

11. (Optional.) Enable the Traffic Flow Confidentiality (TFC) padding feature.

tfc enable

By default, the TFC padding feature is disabled.

Configuring an IKE-based IPsec policy by using an IPsec policy template

1. Enter system view.

system-view

2. Create an IPsec policy template and enter its view.

ipsec { ipv6-policy-template | policy-template } template-name seq-number

3. (Optional.) Configure a description for the IPsec policy template.

description text

By default, no description is configured.

4. (Optional.) Specify an ACL for the IPsec policy template.

security acl [ ipv6 ] { acl-number | name acl-name } [ aggregation | per-host ]

By default, no ACL is specified for an IPsec policy template.

You can specify only one ACL for an IPsec policy template.

5. Specify IPsec transform sets for the IPsec policy template.

transform-set transform-set-name&<1-6>

By default, no IPsec transform sets are specified for an IPsec policy template.

6. Specify an IKE profile or IKEv2 profile for the IPsec policy template.

¡ Specify an IKE profile.

ike-profile profile-name

By default, no IKE profile is specified for an IPsec policy template.

Make sure the specified IKE profile is not used by another IPsec policy or IPsec policy template.

¡ Specify an IKEv2 profile.

ikev2-profile profile-name

By default, no IKEv2 profile is specified for an IPsec policy template.

7. Specify the local IP address of the IPsec tunnel.

local-address { ipv4-address | ipv6 ipv6-address }

The default local IPv4 address and IPv6 address is the primary IPv4 address and first IPv6 address of the interface where the IPsec policy is applied.

The local IP address specified by this command must be the same as the IP address used as the local IKE identity.

In a VRRP network, the local IP address must be the virtual IP address of the VRRP group to which the IPsec-applied interface belongs.

The local address cannot be a secondary IP address of the interface where the IPsec policy is applied.

8. Specify the remote IP address of the IPsec tunnel.

remote-address { [ ipv6 ] host-name | ipv4-address | ipv6 ipv6-address }

By default, the remote IP address of the IPsec tunnel is not specified.

9. (Optional.) Set the lifetime and idle timeout for the IPsec SA.

¡ Set the IPsec SA lifetime.

sa duration { time-based seconds | traffic-based kilobytes }

By default, the global SA lifetime is used.

¡ Set the IPsec SA idle timeout.

sa idle-time seconds

By default, the global IPsec SA idle timeout is used.

10. (Optional.) Enable the Traffic Flow Confidentiality (TFC) padding feature.

tfc enable

By default, the TFC padding feature is disabled.

11. Return to system view.

quit

12. Create an IPsec policy by using the IPsec policy template.

ipsec { ipv6-policy | policy } policy-name seq-number isakmp template template-name

Applying an IPsec policy to an interface

Restrictions and guidelines

An IKE-based IPsec policy that is bound to a source interface can be applied to multiple interfaces.

A manual IPsec policy can be applied to only one interface.

To cancel the IPsec protection, remove the application of the IPsec policy.

Procedure

1. Enter system view.

system-view

2. Enter interface view.

interface interface-type interface-number

3. Apply an IPsec policy to the interface.

ipsec apply { ipv6-policy | policy } policy-name

By default, no IPsec policy is applied to an interface.

On one interface, you can apply only one IPv4 IPsec policy and one IPv6 IPsec policy.

4. Specify a traffic processing slot for the interface.

service slot slot-number

By default, no traffic processing slot is specified for an interface. Traffic on an interface is processed on the slot at which the traffic arrives.

This step is required when the following conditions are met:

¡ An IKE-based IPsec policy is applied to a global logical interface, such as VLAN interface and tunnel interface.

¡ The IPsec anti-replay feature is globally enabled.

Enabling ACL checking for de-encapsulated packets

About this task

This feature compares the de-encapsulated incoming IPsec packets against the ACL in the IPsec policy and discards those that do not match any permit rule of the ACL. This feature can protect networks against attacks using forged IPsec packets.

This feature applies only to tunnel-mode IPsec.

Procedure

1. Enter system view.

system-view

2. Enable ACL checking for de-encapsulated packets.

ipsec decrypt-check enable

By default, ACL checking for de-encapsulated packets is enabled.

Configuring IPsec anti-replay

About this task

IPsec anti-replay protects networks against anti-replay attacks by using a sliding window mechanism called anti-replay window. This feature checks the sequence number of each received IPsec packet against the current IPsec packet sequence number range of the sliding window. If the sequence number is not in the current sequence number range, the packet is considered a replayed packet and is discarded.

IPsec packet de-encapsulation involves complicated calculation. De-encapsulation of replayed packets is not required, and the de-encapsulation process consumes large amounts of resources and degrades performance, resulting in DoS. IPsec anti-replay can check and discard replayed packets before de-encapsulation.

In some situations, service data packets are received in a different order than their original order. The IPsec anti-replay feature drops them as replayed packets, which impacts communications. If this happens, disable IPsec anti-replay checking or adjust the size of the anti-replay window as required.

Restrictions and guidelines

IPsec anti-replay does not affect manually created IPsec SAs. According to the IPsec protocol, only IKE-based IPsec SAs support anti-replay.

Set the anti-replay window size as small as possible to reduce the impact on system performance.

Failure to detect anti-replay attacks might result in denial of services. If you want to disable IPsec anti-replay, make sure you understand the impact of the operation on network security.

Procedure

1. Enter system view.

system-view

2. Enable IPsec anti-replay.

ipsec anti-replay check

By default, IPsec anti-replay is enabled.

3. Set the size of the IPsec anti-replay window.

ipsec anti-replay window width

The default size is 64.

Binding a source interface to an IPsec policy

About this task

For high availability, a core device is usually connected to an ISP through two links, which operate in backup or load sharing mode. The two interfaces negotiate with their peers to establish IPsec SAs respectively. When one interface fails and a link failover occurs, the other interface needs to take some time to renegotiate SAs, resulting in service interruption.

To solve these problems, bind a source interface to an IPsec policy and apply the policy to both interfaces. This enables the two physical interfaces to use the same source interface to negotiate IPsec SAs. As long as the source interface is up, the negotiated IPsec SAs will not be removed and will keep working, regardless of link failover.

Restrictions and guidelines

Only the IKE-based IPsec policies can be bound to a source interface.

An IPsec policy can be bound to only one source interface.

A source interface can be bound to multiple IPsec policies.

If the source interface bound to an IPsec policy is removed, the IPsec policy becomes a common IPsec policy.

If no local address is specified for an IPsec policy that has been bound to a source interface, the IPsec policy uses the IP address of the bound source interface to perform IKE negotiation. If a local address is specified, the IPsec policy uses the local address to perform IKE negotiation.

Procedure

1. Enter system view.

system-view

2. Bind a source interface to an IPsec policy.

ipsec { ipv6-policy | policy } policy-name local-address interface-type interface-number

By default, no source interface is bound to an IPsec policy.

Enabling QoS pre-classify

About this task

When both an IPsec policy and a QoS policy are applied to an interface, QoS classifies packets by using the new headers added by IPsec. If you want QoS to classify packets by using the headers of the original IP packets, enable the QoS pre-classify feature.

Restrictions and guidelines

If you configure both IPsec and QoS on an interface, make sure the IPsec traffic classification rules match the QoS traffic classification rules. If the rules do not match, QoS might classify the packets of one IPsec SA to different queues, causing packets to be sent out of order. When IPsec anti-replay is enabled, IPsec will drop the incoming packets that are out of the anti-replay window, resulting in packet loss.

IPsec traffic classification rules are determined by the rules of the specified ACL. For more information about QoS policy and classification, see QoS Configuration Guide.

Procedure

1. Enter system view.

system-view

2. Enter IPsec policy view or IPsec policy template view.

¡ Enter IPsec policy view.

ipsec { ipv6-policy | policy } policy-name seq-number [ isakmp | manual ]

¡ Enter IPsec policy template view.

ipsec { ipv6-policy-template | policy-template } template-name seq-number

3. Enable QoS pre-classify.

qos pre-classify

By default, QoS pre-classify is disabled.

Configuring the DF bit of IPsec packets

About this task

Perform this task to configure the Don't Fragment (DF) bit in the new IP header of IPsec packets in one of the following ways:

· clear—Clears the DF bit in the new header.

· set—Sets the DF bit in the new header.

· copy—Copies the DF bit in the original IP header to the new IP header.

You can configure the DF bit in system view and interface view. The interface-view DF bit setting takes precedence over the system-view DF bit setting. If the interface-view DF bit setting is not configured, the interface uses the system-view DF bit setting.

Restrictions and guidelines for DF bit configuration for IPsec packets

The DF bit setting takes effect only in tunnel mode, and it changes the DF bit in the new IP header rather than the original IP header.

Configure the same DF bit setting on the interfaces where the same IPsec policy bound to a source interface is applied.

If the DF bit is set, the devices on the path cannot fragment the IPsec packets. To prevent IPsec packets from being discarded, make sure the path MTU is larger than the IPsec packet size. As a best practice, clear the DF bit if you cannot make sure the path MTU is larger than the IPsec packet size.

Configuring the DF bit of IPsec packets on an interface

1. Enter system view.

system-view

2. Enter interface view.

interface interface-type interface-number

3. Configure the DF bit of IPsec packets on the interface.

ipsec df-bit { clear | copy | set }

By default, the interface uses the global DF bit setting.

Configuring the DF bit of IPsec packets globally

1. Enter system view.

system-view

2. Configure the DF bit of IPsec packets globally.

ipsec global-df-bit { clear | copy | set }

By default, IPsec copies the DF bit in the original IP header to the new IP header.

Configuring IPsec RRI

Restrictions and guidelines

Enabling IPsec RRI for an IPsec policy deletes all existing IPsec SAs created by this IPsec policy. IPsec RRI creates static routes according to new IPsec SAs.

Disabling IPsec RRI for an IPsec policy deletes all existing IPsec SAs created by this IPsec policy and the associated static routes.

IPsec RRI is supported in both tunnel mode and transport mode.

If you change the preference value or tag value for an IPsec policy, the device deletes all IPsec SAs created by this IPsec policy, and the associated static routes. The change takes effect for future IPsec RRI-created static routes.

IPsec RRI does not generate a static route to a destination address to be protected if the destination address is not defined in the ACL used by an IPsec policy or an IPsec policy template. You must manually configure a static route to the destination address.

Procedure

1. Enter system view.

system-view

2. Enter IPsec policy view or IPsec policy template view.

¡ Enter IPsec policy view.

ipsec { policy | ipv6-policy } policy-name seq-number isakmp

¡ Enter IPsec policy template view.

ipsec { ipv6-policy-template | policy-template } template-name seq-number

3. Enable IPsec RRI.

reverse-route [ next-hop [ ipv6 ] ip-address ] dynamic

By default, IPsec RRI is disabled.

4. (Optional.) Set the preference value for the static routes created by IPsec RRI.

reverse-route preference number

The default value is 60.

5. (Optional.) Set the tag value for the static routes created by IPsec RRI.

reverse-route tag tag-value

The default value is 0.

Configuring the global IPsec SA lifetime and idle timeout

About this task

If the IPsec SA lifetime and idle timeout are not configured in an IPsec policy, IPsec policy template, or IPsec profile, the global settings are used.

When IKE negotiates IPsec SAs, it uses the local lifetime settings or those proposed by the peer, whichever are smaller.

An IPsec SA can have both a time-based lifetime and a traffic-based lifetime. The IPsec SA expires when either lifetime expires.

Procedure

1. Enter system view.

system-view

2. Set the global IPsec SA lifetime, soft lifetime buffer, or idle timeout.

¡ Set the global IPsec SA lifetime.

ipsec sa global-duration { time-based seconds | traffic-based kilobytes }

By default, the time-based SA lifetime is 3600 seconds, and the traffic-based SA lifetime is 1843200 kilobytes.

¡ Set the global time-based or traffic-based IPsec SA soft lifetime buffer.

ipsec sa global-soft-duration buffer { time-based seconds | traffic-based kilobytes }

By default, no global IPsec SA soft lifetime buffers are configured.

¡ Set the global SA idle timeout.

ipsec sa idle-time seconds

By default, the global IPsec SA idle timeout feature is disabled.

Configuring IPsec fragmentation

About this task

Perform this task to configure the device to fragment packets before or after IPsec encapsulation.

If you configure the device to fragment packets before IPsec encapsulation, the device predetermines the encapsulated packet size before the actual encapsulation. If the encapsulated packet size exceeds the MTU of the output interface, the device fragments the packets before encapsulation. If a packet's DF bit is set, the device drops the packet and sends an ICMP error message.

If you configure the device to fragment packets after IPsec encapsulation, the device directly encapsulates the packets and fragments the encapsulated packets in subsequent service modules.

Restrictions and guidelines

This feature takes effect on IPsec protected IPv4 packets.

Procedure

1. Enter system view.

system-view

2. Configure IPsec fragmentation.

ipsec fragmentation { after-encryption | before-encryption }

By default, the device fragments packets before IPsec encapsulation.

Setting the maximum number of IPsec tunnels

Restrictions and guidelines

To maximize concurrent performance of IPsec when memory is sufficient, increase the maximum number of IPsec tunnels. To ensure service availability when memory is insufficient, decrease the maximum number of IPsec tunnels.

Procedure

1. Enter system view.

system-view

2. Set the maximum number of IPsec tunnels.

ipsec limit max-tunnel tunnel-limit

By default, the number of IPsec tunnels is not limited.

Enabling logging for IPsec packets

About this task

Perform this task to enable logging for IPsec packets that are discarded for reasons such as IPsec SA lookup failure, AH-ESP authentication failure, and ESP encryption failure. The log information includes the source and destination IP addresses, SPI value, and sequence number of a discarded IPsec packet, and the reason for the discard.

Procedure

1. Enter system view.

system-view

2. Enable logging for IPsec packets.

ipsec logging packet enable

By default, logging for IPsec packets is disabled.

Enabling logging for IPsec negotiation

About this task

This feature enables the device to output logs for the IPsec negotiation process.

Procedure

1. Enter system view.

system-view

2. Enable logging for IPsec negotiation.

ipsec logging negotiation enable

By default, logging for IPsec negotiation is disabled.

Configuring SNMP notifications for IPsec

About this task

After you enable SNMP notifications for IPsec, the IPsec module notifies the NMS of important module events. The notifications are sent to the device's SNMP module. For the notifications to be sent correctly, you must also configure SNMP on the device. For more information about SNMP notifications, see Network Management and Monitoring Configuration Guide.

To generate and output SNMP notifications for a specific IPsec failure or event type, perform the following tasks:

1. Enable SNMP notifications for IPsec globally.

2. Enable SNMP notifications for the failure or event type.

Procedure

1. Enter system view.

system-view

2. Enable SNMP notifications for IPsec globally.

snmp-agent trap enable ipsec global

By default, SNMP notifications for IPsec are disabled.

3. Enable SNMP notifications for the specified failure or event types.

By default, SNMP notifications for all failure and event types are disabled.

Display and maintenance commands for IPsec

Execute display commands in any view and reset commands in user view.

Task	Command
Display IPsec policy information.	display ipsec { ipv6-policy \| policy } [ policy-name [ seq-number ] ]
Display IPsec policy template information.	display ipsec { ipv6-policy-template \| policy-template } [ template-name [ seq-number ] ]
Display IPsec SA information.	display ipsec sa [ brief \| count \| interface interface-type interface-number \| { ipv6-policy \| policy } policy-name [ seq-number ] \| remote [ ipv6 ] ip-address ]
Display IPsec statistics.	display ipsec statistics [ tunnel-id tunnel-id ]
Display IPsec transform set information.	display ipsec transform-set [ transform-set-name ]
Display IPsec tunnel information.	display ipsec tunnel { brief \| count \| tunnel-id tunnel-id }
Clear IPsec SAs.	reset ipsec sa [ { ipv6-policy \| policy } policy-name [ seq-number ] \| remote { ipv4-address \| ipv6 ipv6-address } \| spi { ipv4-address \| ipv6 ipv6-address } { ah \| esp } spi-num ]
Clear IPsec statistics.	reset ipsec statistics [ tunnel-id tunnel-id ]

Configuring IKE

Unless otherwise specified, the term "IKE" in this chapter refers to IKEv1.

About IKE

Built on a framework defined by ISAKMP, Internet Key Exchange (IKE) provides automatic key negotiation and SA establishment services for IPsec.

Benefits of IKE

IKE provides the following benefits for IPsec:

· Automatically negotiates IPsec parameters.

· Performs DH exchanges to calculate shared keys, making sure each SA has a key that is independent of other keys.

· Automatically negotiates SAs when the sequence number in the AH or ESP header overflows, making sure IPsec can provide the anti-replay service by using the sequence number.

Relationship between IPsec and IKE

As shown in Figure 7, IKE negotiates SAs for IPsec and transfers the SAs to IPsec, and IPsec uses the SAs to protect IP packets.

Figure 7 Relationship between IKE and IPsec

IKE negotiation process

IKE negotiates keys and SAs for IPsec in two phases:

1. Phase 1—The two peers establish an IKE SA, a secure, authenticated channel for communication.

2. Phase 2—Using the IKE SA established in phase 1, the two peers negotiate to establish IPsec SAs.

Phase 1 negotiation can use the main mode or aggressive mode.

IKE exchange process in main mode

As shown in Figure 8, the main mode of IKE negotiation in phase 1 involves three pairs of messages:

· SA exchange—Used for negotiating the IKE security policy.

· Key exchange—Used for exchanging the DH public value and other values, such as the random number. The two peers use the exchanged data to generate key data and use the encryption key and authentication key to ensure the security of IP packets.

· ID and authentication data exchange—Used for identity authentication.

Figure 8 IKE exchange process in main mode

IKE exchange process in aggressive mode

As shown in Figure 9, the process of phase 1 IKE negotiation in aggressive mode is as follows:

1. The initiator (peer 1) sends a message containing the local IKE information to peer 2. The message includes parameters used for IKE SA establishment, keying data, and peer 1's identity information.

2. Peer 2 chooses the IKE establishment parameters to use, generate the key, and authenticate peer 1's identity. Then it sends the IKE data to peer 1.

3. Peer 1 generates the key, authenticates peer 2's identity, and sends the results to peer 1.

After the preceding process, an IKE SA is established between peer 1 and peer 2.

The aggressive mode is faster than the main mode but it does not provide identity information protection. The main mode provides identity information protection but is slower. Choose the appropriate negotiation mode according to your requirements.

Figure 9 IKE exchange process in aggressive mode

IKE security mechanism

IKE has a series of self-protection mechanisms and supports secure identity authentication, key distribution, and IPsec SA establishment on insecure networks.

Identity authentication

The IKE identity authentication mechanism is used to authenticate the identity of the communicating peers. The device supports the following identity authentication methods:

· Preshared key authentication—Two communicating peers use the pre-configured shared key for identity authentication.

· RSA signature authentication and DSA signature authentication—Two communicating peers use the digital certificates issued by the CA for identity authentication.

The preshared key authentication method does not require certificates and is easy to configure. It is usually deployed in small networks.

The signature authentication methods provide higher security and are usually deployed in networks with the headquarters and some branches. When deployed in a network with many branches, a signature authentication method can simplify the configuration because only one PKI domain is required. If you use the preshared key authentication method, you must configure a preshared key for each branch on the Headquarters node.

DH algorithm

The DH algorithm is a public key algorithm. With this algorithm, two peers can exchange keying material and then use the material to calculate the shared keys. Due to the decryption complexity, a third party cannot decrypt the keys even after intercepting all keying materials.

PFS

The Perfect Forward Secrecy (PFS) feature is a security feature based on the DH algorithm. After PFS is enabled, an additional DH exchange is performed in IKE phase 2 to make sure IPsec keys have no derivative relations with IKE keys and a broken key brings no threats to other keys.

Protocols and standards

· RFC 2408, Internet Security Association and Key Management Protocol (ISAKMP)

· RFC 2409, The Internet Key Exchange (IKE)

· RFC 2412, The OAKLEY Key Determination Protocol

· Internet Draft, draft-ietf-ipsec-isakmp-xauth-06

Internet Draft, draft-dukes-ike-mode-cfg-02

IKE tasks at a glance

To configure IKE, perform the following tasks:

1. (Optional.) Configuring an IKE profile

a. Creating an IKE profile

b. Configuring peer IDs for the IKE profile

c. Specifying the IKE keychain or PKI domain

d. Configuring the IKE phase 1 negotiation mode

e. Specifying IKE proposals for the IKE profile

f. Configuring the local ID for the IKE profile

g. Configuring optional features for the IKE profile

2. Configuring an IKE proposal

3. Configuring an IKE keychain

4. (Optional.) Configuring the global identity information

5. (Optional.) Configuring the IKE keepalive feature

6. (Optional.) Configuring the IKE NAT keepalive feature

7. (Optional.) Configuring global IKE DPD

8. (Optional.) Enabling invalid SPI recovery

9. (Optional.) Setting the maximum number of IKE SAs

10. (Optional.) Configuring an IKE IPv4 address pool

11. (Optional.) Configuring SNMP notifications and logging for IKE

Prerequisites for IKE configuration

Determine the following parameters prior to IKE configuration:

· The algorithms to be used during IKE negotiation, including the identity authentication method, encryption algorithm, authentication algorithm, and DH group.

¡ Different algorithms provide different levels of protection. A stronger algorithm provides more resistance to decryption but uses more resources.

¡ A DH group that uses more bits provides higher security but needs more time for processing.

· The preshared key or PKI domain for IKE negotiation. For more information about PKI, see "Configuring PKI."

· The IKE-based IPsec policies for the communicating peers. If you do not specify an IKE profile in an IPsec policy, the device selects an IKE profile for the IPsec policy. If no IKE profile is configured, the globally configured IKE settings are used. For more information about IPsec, see "Configuring IPsec."

Configuring an IKE profile

Creating an IKE profile

About this task

Perform this task to create an IKE profile.

An IKE profile is intended to provide a set of parameters for IKE negotiation.

Procedure

1. Enter system view.

system-view

2. Create an IKE profile and enter its view.

ike profile profile-name

Configuring peer IDs for the IKE profile

About this task

Perform this task to configure the peer IDs for IKE profile matching. When the device needs to select an IKE profile for IKE negotiation with a peer, it compares the received peer ID with the peer IDs of its local IKE profiles. If a match is found, it uses the IKE profile with the matching peer ID for IKE negotiation.

Restrictions and guidelines

For an IKE profile, you can configure multiple peer IDs. A peer ID configured earlier has a higher priority.

Two IKE peers must both have or both not have peer IDs configured.

Procedure

1. Enter system view.

system-view

2. Enter IKE profile view.

ike profile profile-name

3. Configure a peer ID for the IKE profile.

match remote { certificate policy-name | identity { address { { ipv4-address [ mask | mask-length ] | range low-ipv4-address high-ipv4-address } | ipv6 { ipv6-address [ prefix-length ] | range low-ipv6-address high-ipv6-address } } | fqdn fqdn-name | user-fqdn user-fqdn-name } }

Specifying the IKE keychain or PKI domain

Restrictions and guidelines

Configure the IKE keychain or PKI domain for the IKE proposals to use. To use digital signature authentication, configure a PKI domain. To use preshared key authentication, configure an IKE keychain.

Procedure

1. Enter system view.

system-view

2. Enter IKE profile view.

ike profile profile-name

3. Specify the keychain for preshared key authentication or the PKI domain used to request a certificate for digital signature authentication.

¡ Specify the keychain.

keychain keychain-name

¡ Specify the PKI domain.

certificate domain domain-name

By default, no IKE keychain or PKI domain is specified in an IKE profile.

Configuring the IKE phase 1 negotiation mode

Restrictions and guidelines

Specify the IKE phase 1 negotiation mode (main or aggressive) that the device uses as the initiator. When the device acts as the responder, it uses the IKE negotiation mode of the initiator.

Procedure

1. Enter system view.

system-view

2. Enter IKE profile view.

ike profile profile-name

3. Specify the IKE negotiation mode for phase 1.

exchange-mode { aggressive | main }

By default, IKE negotiation in phase 1 uses the main mode.

Specifying IKE proposals for the IKE profile

Restrictions and guidelines

Specify the IKE proposals that the device can use as the initiator. An IKE proposal specified earlier has a higher priority. When the device acts as the responder, it uses the IKE proposals configured in system view to match the IKE proposals received from the initiator. If no matching proposal is found, the negotiation fails.

Procedure

1. Enter system view.

system-view

2. Enter IKE profile view.

ike profile profile-name

3. Specify IKE proposals for the IKE profile.

proposal proposal-number&<1-6>

By default, no IKE proposals are specified for an IKE profile and the IKE proposals configured in system view are used for IKE negotiation.

Configuring the local ID for the IKE profile

Restrictions and guidelines

For digital signature authentication, the device can use an ID of any type. If the local ID is an IP address that is different from the IP address in the local certificate, the device uses the FQDN (the device name configured by using the sysname command) instead.

For preshared key authentication, the device can use an ID of any type other than the DN.

Procedure

1. Enter system view.

system-view

2. Enter IKE profile view.

ike profile profile-name

3. Configure the local ID.

local-identity { address { ipv4-address | ipv6 ipv6-address } | dn | fqdn [ fqdn-name ] | user-fqdn [ user-fqdn-name ] }

By default, no local ID is configured for an IKE profile, and an IKE profile uses the local ID configured in system view. If the local ID is not configured in system view, the IKE profile uses the IP address of the interface to which the IPsec policy or IPsec policy template is applied as the local ID.

Configuring optional features for the IKE profile

1. Enter system view.

system-view

2. Enter IKE profile view.

ike profile profile-name

3. Configure optional features as needed.

¡ Configure IKE DPD.

dpd interval interval [ retry seconds ] { on-demand | periodic }

By default, IKE DPD is not configured for an IKE profile and an IKE profile uses the DPD settings configured in system view. If IKE DPD is not configured in system view either, the device does not perform dead IKE peer detection.

The IKE DPD settings configured in the IKE profile view take precedence over those configured in system view.

¡ Specify the local interface or IP address to which the IKE profile can be applied.

match local address { interface-type interface-number | { ipv4-address | ipv6 ipv6-address } }

By default, an IKE profile can be applied to any local interface or IP address.

An IKE profile configured with this command has a higher priority over those not configured with this command.

¡ Specify a priority for the IKE profile.

priority priority

By default, the priority of an IKE profile is 100.

The device selects a local IKE profile for IKE negotiation as follows:

- First, it selects an IKE profile with the match local address command configured.

- If a tie exists, it selects the IKE profile with a smaller priority number.

- If a tie still exists, it selects the IKE profile configured earlier.

¡ Enable client authentication.

client-authentication xauth

By default, client authentication is disabled.

This feature is typically used on the IPsec gateway at the enterprise center. It enables the IPsec gateway to perform extended (XAUTH) authentication on remote users through AAA after IKE phase 1 negotiation. AAA configuration is also required on the IPsec gateway for client authentication.

¡ Enable AAA authorization.

aaa authorization domain domain-name username user-name

By default, AAA authorization is disabled.

The AAA authorization feature enables IKE to request authorization attributes, such as the IKE IPv4 address pool, from AAA. IKE uses the address pool to assign IPv4 addresses to remote users. For more information about AAA authorization, see AAA in User Access and Authentication Configuration Guide.

¡ (Optional.) Set the IKE SA soft lifetime buffer time.

sa soft-duration buffer seconds

By default, the IKE SA soft lifetime buffer time is not configured.

Set the IKE SA soft lifetime buffer time to determine the IKE SA soft lifetime. The SA soft lifetime is calculated as follows: SA soft lifetime = SA lifetime – SA soft lifetime buffer. A new IKE SA will be negotiated when the SA soft lifetime timer expires.

Configuring an IKE proposal

About this task

An IKE proposal defines a set of attributes describing how IKE negotiation in phase 1 should take place. You can create multiple IKE proposals with different priorities. The priority of an IKE proposal is represented by its sequence number. The lower the sequence number, the higher the priority.

Two peers must have at least one matching IKE proposal for successful IKE negotiation. During IKE negotiation:

· The initiator sends its IKE proposals to the peer.

¡ If the initiator is using an IPsec policy with an IKE profile, the initiator sends all IKE proposals specified in the IKE profile to the peer. An IKE proposal specified earlier for the IKE profile has a higher priority.

¡ If the initiator is using an IPsec policy with no IKE profile, the initiator sends all its IKE proposals to the peer. An IKE proposal with a smaller number has a higher priority.

· The peer searches its own IKE proposals for a match. The search starts from the IKE proposal with the highest priority and proceeds in descending order of priority until a match is found. The matching IKE proposals are used to establish the IKE SA. If all user-defined IKE proposals are found mismatching, the two peers use their default IKE proposals to establish the IKE SA.

Two matching IKE proposals have the same encryption algorithm, authentication method, authentication algorithm, and DH group. The SA lifetime takes the smaller one of the two proposals' SA lifetime settings.

Procedure

1. Enter system view.

system-view

2. Create an IKE proposal and enter its view.

ike proposal proposal-number

By default, a default IKE proposal exists.

3. Configure a description for the IKE proposal.

description

By default, an IKE proposal does not have a description.

4. Specify an encryption algorithm for the IKE proposal.

encryption-algorithm { 3des-cbc | aes-cbc-128 | aes-cbc-192 | aes-cbc-256 | des-cbc }

By default, the 56-bit DES encryption algorithm in CBC mode is used.

5. Specify an authentication method for the IKE proposal.

authentication-method { dsa-signature | pre-share | rsa-de | rsa-signature }

By default, the preshared key authentication method is used.

6. Specify an authentication algorithm for the IKE proposal.

authentication-algorithm { md5 | sha | sha256 | sha384 | sha512 }

By default, the HMAC-SHA1 authentication algorithm is used.

7. Specify a DH group for key negotiation in phase 1.

dh { group1 | group14 | group2 | group24 | group5 }

DH group 1 (the 768-bit DH group) is used by default.

8. (Optional.) Set the IKE SA lifetime for the IKE proposal.

sa duration seconds

By default, the IKE SA lifetime is 86400 seconds.

If the IPsec SA lifetime is also configured, set the IKE SA lifetime longer than the IPsec SA lifetime as a best practice.

Configuring an IKE keychain

About this task

Perform this task when you configure the IKE to use the preshared key for authentication.

Follow these guidelines when you configure an IKE keychain:

· Two peers must be configured with the same preshared key to pass preshared key authentication.

· You can specify the local address configured in IPsec policy or IPsec policy template view (using the local-address command) for the IKE keychain to be applied. If no local address is configured, specify the IP address of the interface that uses the IPsec policy.

· The device determines the priority of an IKE keychain as follows:

a. The device examines the existence of the match local address command. An IKE keychain with the match local address command configured has a higher priority.

b. If a tie exists, the device compares the priority numbers. An IKE keychain with a smaller priority number has a higher priority.

c. If a tie still exists, the device prefers an IKE keychain configured earlier.

Procedure

1. Enter system view.

system-view

2. Create an IKE keychain and enter its view.

ike keychain keychain-name

3. Configure a preshared key.

pre-shared-key { address { ipv4-address [ mask | mask-length ] | ipv6 ipv6-address [ prefix-length ] } | hostname host-name } key { cipher | simple } string

By default, no preshared key is configured.

4. (Optional.) Specify a local interface or IP address to which the IKE keychain can be applied.

match local address { interface-type interface-number | { ipv4-address | ipv6 ipv6-address } }

By default, an IKE keychain can be applied to any local interface or IP address.

5. (Optional.) Specify a priority for the IKE keychain.

priority priority

The default priority is 100.

Configuring the global identity information

Restrictions and guidelines

The global identity can be used by the device for all IKE SA negotiations, and the local identity (set by the local-identity command) can be used only by the device that uses the IKE profile.

When signature authentication is used, you can set any type of the identity information.

When preshared key authentication is used, you cannot set the DN as the identity.

Procedure

1. Enter system view.

system-view

2. Configure the global identity to be used by the local end.

ike identity { address { ipv4-address | ipv6 ipv6-address }| dn | fqdn [ fqdn-name ] | user-fqdn [ user-fqdn-name ] }

By default, the IP address of the interface to which the IPsec policy or IPsec policy template is applied is used as the IKE identity.

3. (Optional.) Configure the local device to always obtain the identity information from the local certificate for signature authentication.

ike signature-identity from-certificate

By default, the local end uses the identity information specified by local-identity or ike identity for signature authentication.

Configure this command when the aggressive mode and signature authentication are used and the device interconnects with a Comware 5-based peer device. Comware 5 supports only DN for signature authentication.

Configuring the IKE keepalive feature

About this task

IKE sends keepalive packets to query the liveness of the peer. If the peer is configured with the keepalive timeout time, you must configure the keepalive interval on the local device. If the peer receives no keepalive packets during the timeout time, the IKE SA is deleted along with the IPsec SAs it negotiated.

Restrictions and guidelines

Configure IKE DPD instead of IKE keepalive unless IKE DPD is not supported on the peer. The IKE keepalive feature sends keepalives at regular intervals, which consumes network bandwidth and resources.

The keepalive timeout time configured on the local device must be longer than the keepalive interval configured at the peer. Since it seldom occurs that more than three consecutive packets are lost on a network, you can set the keepalive timeout three times as long as the keepalive interval.

Procedure

1. Enter system view.

system-view

2. Set the IKE SA keepalive interval.

ike keepalive interval interval

By default, no keepalives are sent to the peer.

3. Set the IKE SA keepalive timeout time.

ike keepalive timeout seconds

By default, IKE SA keepalive never times out.

Configuring the IKE NAT keepalive feature

About this task

If IPsec traffic passes through a NAT device, you must configure the NAT traversal feature. If no packet travels across an IPsec tunnel in a period of time, the NAT sessions are aged and deleted, disabling the tunnel from transmitting data to the intended end. To prevent NAT sessions from being aged, configure the NAT keepalive feature on the IKE gateway behind the NAT device to send NAT keepalive packets to its peer periodically to keep the NAT session alive.

Procedure

1. Enter system view.

system-view

2. Set the IKE NAT keepalive interval.

ike nat-keepalive seconds

The default interval is 20 seconds.

Configuring global IKE DPD

About this task

DPD detects dead peers. It can operate in periodic mode or on-demand mode.

· Periodic DPD—Sends a DPD message at regular intervals. It features an earlier detection of dead peers, but consumes more bandwidth and CPU.

· On-demand DPD—Sends a DPD message based on traffic. When the device has traffic to send and is not aware of the liveness of the peer, it sends a DPD message to query the status of the peer. If the device has no traffic to send, it never sends DPD messages. As a best practice, use the on-demand mode.

The IKE DPD works as follows:

1. The local device sends a DPD message to the peer, and waits for a response from the peer.

2. If the peer does not respond within the retry interval specified by the retry seconds parameter, the local device resends the message.

3. If still no response is received within the retry interval, the local end sends the DPD message again. The system allows a maximum of four retries.

4. If the local device receives no response after four retries, the device considers the peer to be dead, and deletes the IKE SA along with the IPsec SAs it negotiated.

5. If the local device receives a response from the peer during the detection process, the peer is considered alive. The local device performs a DPD detection again when the triggering interval is reached or it has traffic to send, depending on the DPD mode.

Restrictions and guidelines

When DPD settings are configured in both IKE profile view and system view, the DPD settings in IKE profile view apply. If DPD is not configured in IKE profile view, the DPD settings in system view apply.

It is a good practice to set the triggering interval longer than the retry interval so that a DPD detection is not triggered during a DPD retry.

Procedure

1. Enter system view.

system-view

2. Enable sending IKE DPD messages.

ike dpd interval interval [ retry seconds ] { on-demand | periodic }

By default, IKE DPD is disabled.

Enabling invalid SPI recovery

About this task

An IPsec "black hole" occurs when one IPsec peer fails (for example, a peer can fail if a reboot occurs). One peer fails and loses its SAs with the other peer. When an IPsec peer receives a data packet for which it cannot find an SA, an invalid SPI is encountered. The peer drops the data packet and tries to send an SPI invalid notification to the data originator. This notification is sent by using the IKE SA. Because no IKE SA is available, the notification is not sent. The originating peer continues sending the data by using the IPsec SA that has the invalid SPI, and the receiving peer keeps dropping the traffic.

The invalid SPI recovery feature enables the receiving peer to set up an IKE SA with the originator so that an SPI invalid notification can be sent. Upon receiving the notification, the originating peer deletes the IPsec SA that has the invalid SPI. If the originator has data to send, new SAs will be set up.

Restrictions and guidelines

Use caution when you enable the invalid SPI recovery feature because using this feature can result in a DoS attack. Attackers can make a great number of invalid SPI notifications to the same peer.

Procedure

1. Enter system view.

system-view

2. Enable invalid SPI recovery.

ike invalid-spi-recovery enable

By default, the invalid SPI recovery is disabled.

Setting the maximum number of IKE SAs

About this task

You can set the maximum number of half-open IKE SAs and the maximum number of established IKE SAs.

· The supported maximum number of half-open IKE SAs depends on the device's processing capability. Adjust the maximum number of half-open IKE SAs to make full use of the device's processing capability without affecting the IKE SA negotiation efficiency.

· The supported maximum number of established IKE SAs depends on the device's memory space. Adjust the maximum number of established IKE SAs to make full use of the device's memory space without affecting other applications in the system.

Procedure

1. Enter system view.

system-view

2. Set the maximum number of half-open IKE SAs and the maximum number of established IKE SAs.

ike limit { max-negotiating-sa negotiation-limit | max-sa sa-limit }

By default, the maximum number of half-open IKE SAs and IPsec SAs is 200, and there is no limit to the maximum number of IKE SAs.

Configuring an IKE IPv4 address pool

About this task

To perform centralized management on remote users, an IPsec gateway can use an IPv4 address pool to assign private IPv4 addresses to remote users.

You must use an IKE IPv4 address pool together with AAA authorization by specifying the IKE IPv4 address pool as an AAA authorization attribute. For more information about AAA authorization, see AAA in User Access and Authentication Configuration Guide.

Procedure

1. Enter system view.

system-view

2. Configure an IKE IPv4 address pool.

ike address-group group-name start-ipv4-address end-ipv4-address [ mask | mask-length ]

Configuring SNMP notifications and logging for IKE

Configuring SNMP notifications for IKE

About this task

After you enable SNMP notifications for IKE, the IKE module notifies the NMS of important module events. The notifications are sent to the device's SNMP module. For the notifications to be sent correctly, you must also configure SNMP on the device. For more information about SNMP notifications, see Network Management and Monitoring Configuration Guide.

To generate and output SNMP notifications for a specific IKE failure or event type, perform the following tasks:

1. Enable SNMP notifications for IKE globally.

2. Enable SNMP notifications for the failure or event type.

Procedure

1. Enter system view

system-view

2. Enable SNMP notifications for IKE globally.

snmp-agent trap enable ike global

By default, SNMP notifications for IKE are disabled.

3. Enable SNMP notifications for the specified failure or event types.

By default, SNMP notifications for all failure and event types are disabled.

Enabling logging for IKE negotiation

About this task

This feature enables the device to output logs for the IKE negotiation process.

Procedure

1. Enter system view.

system-view

2. Enable logging for IKE negotiation.

ike logging negotiation enable

By default, logging for IKE negotiation is disabled.

Display and maintenance commands for IKE

Execute display commands in any view and reset commands in user view.

Task	Command
Display configuration information about all IKE proposals.	display ike proposal
Display information about the current IKE SAs.	display ike sa [ verbose [ connection-id connection-id \| remote-address [ ipv6 ] remote-address ] ]
Display IKE statistics.	display ike statistics
Delete IKE SAs.	reset ike sa [ connection-id connection-id ]
Clear IKE MIB statistics.	reset ike statistics

Troubleshooting IKE

IKE negotiation failed because no matching IKE proposals were found

Symptom

1. The IKE SA is in Unknown state.

<Sysname> display ike sa

Connection-ID Remote Flag DOI

------------------------------------------------------------------

1 192.168.222.5 Unknown IPsec

Flags:

RD--READY RL--REPLACED FD-FADING RK-REKEY

2. When IKE event debugging and packet debugging are enabled, the following messages appear:

IKE event debugging message:

The attributes are unacceptable.

IKE packet debugging message:

Construct notification packet: NO_PROPOSAL_CHOSEN.

Analysis

Certain IKE proposal settings are incorrect.

Solution

1. Examine the IKE proposal configuration to see whether the two ends have matching IKE proposals.

2. Modify the IKE proposal configuration to make sure the two ends have matching IKE proposals.

3. If the problem persists, contact H3C Support.

IKE negotiation failed because no IKE proposals or IKE keychains are specified correctly

Symptom

1. The IKE SA is in Unknown state.

<Sysname> display ike sa

Connection-ID Remote Flag DOI

------------------------------------------------------------------

1 192.168.222.5 Unknown IPsec

Flags:

RD--READY RL--REPLACED FD-FADING RK-REKEY

2. The following IKE event debugging or packet debugging message appeared:

IKE event debugging message:

Notification PAYLOAD_MALFORMED is received.

IKE packet debugging message:

Construct notification packet: PAYLOAD_MALFORMED.

Analysis

· If the following debugging information appeared, the matched IKE profile is not using the matched IKE proposal:

Failed to find proposal 1 in profile profile1.

· If the following debugging information appeared, the matched IKE profile is not using the matched IKE keychain:

Failed to find keychain keychain1 in profile profile1.

Solution

1. Verify that the matched IKE proposal (IKE proposal 1 in this debugging message example) is specified for the IKE profile (IKE profile 1 in the example).

2. Verify that the matched IKE keychain (IKE keychain 1 in this debugging message example) is specified for the IKE profile (IKE profile 1 in the example).

3. If the problem persists, contact H3C Support.

IPsec SA negotiation failed because no matching IPsec transform sets were found

Symptom

1. The display ike sa command shows that the IKE SA negotiation succeeded and the IKE SA is in RD state, but the display ipsec sa command shows that the expected IPsec SA has not been negotiated yet.

2. The following IKE debugging message appeared:

The attributes are unacceptable.

Or:

Construct notification packet: NO_PROPOSAL_CHOSEN.

Analysis

Certain IPsec policy settings are incorrect.

Solution

1. Examine the IPsec configuration to see whether the two ends have matching IPsec transform sets.

2. Modify the IPsec configuration to make sure the two ends have matching IPsec transform sets.

3. If the problem persists, contact H3C Support.

IPsec SA negotiation failed due to invalid identity information

Symptom

1. The display ike sa command shows that the IKE SA negotiation succeeded and the IKE SA is in RD state, but the display ipsec sa command shows that the expected IPsec SA has not been negotiated yet.

2. The following IKE debugging message appeared:

Notification INVALID_ID_INFORMATION is received.

Or:

Failed to get IPsec policy when renegotiating IPsec SA. Delete IPsec SA.

Construct notification packet: INVALID_ID_INFORMATION.

Analysis

Certain IPsec policy settings of the responder are incorrect. Verify the settings as follows:

1. Use the display ike sa verbose command to verify that matching IKE profiles were found in IKE negotiation phase 1. If no matching IKE profiles were found and the IPsec policy is using an IKE profile, the IPsec SA negotiation fails.

# Identify whether matching IKE profiles were found in IKE negotiation phase 1. The following output shows that no matching IKE profile was found.

<Sysname> display ike sa verbose

-----------------------------------------------

Connection ID: 3

Outside VPN:

Inside VPN:

Profile:

Transmitting entity: Responder

Initiator cookie: 1bcf453f0a217259

Responder cookie: 5e32a74dfa66a0a4

-----------------------------------------------

Local IP: 192.168.222.5

Local ID type: IPV4_ADDR

Local ID: 192.168.222.5

Remote IP: 192.168.222.71

Remote ID type: IPV4_ADDR

Remote ID: 192.168.222.71

Authentication-method: PRE-SHARED-KEY

Authentication-algorithm: MD5

Encryption-algorithm: 3DES-CBC

Life duration(sec): 86400

Remaining key duration(sec): 85847

Exchange-mode: Main

Diffie-Hellman group: Group 1

NAT traversal: Not detected

Vendor ID index: 0xa1d

Vendor ID sequence number: 0x0

# Identify whether the IPsec policy is using an IKE profile. The following output shows that an IKE profile was used in the IPsec policy.

[Sysname] display ipsec policy

-------------------------------------------

IPsec Policy: policy1

Interface: GigabitEthernet1/0/1

-------------------------------------------

-----------------------------

Sequence number: 1

Mode: ISAKMP

-----------------------------

Description:

Security data flow: 3000

Selector mode: aggregation

Local address: 192.168.222.5

Remote address: 192.168.222.71

Transform set: transform1

IKE profile: profile1

smart-link policy:

SA trigger mode: Auto

SA duration(time based): 3600 seconds

SA duration(traffic based): 1843200 kilobytes

SA soft-duration buffer(time based): 1000 seconds

SA soft-duration buffer(traffic based): 43200 kilobytes

SA idle time: 100 seconds

2. Verify that the ACL specified for the IPsec policy is correctly configured. If the flow range defined by the responder's ACL is smaller than that defined by the initiator's ACL, IPsec proposal matching will fail.

For example, if the initiator's ACL defines a flow from one network segment to another but the responder's ACL defines a flow from one host to another host, IPsec proposal matching will fail.

# On the initiator:

[Sysname] display acl 3000

Advanced IPv4 ACL 3000, 1 rule,

ACL's step is 5

rule 0 permit ip source 192.168.222.0 0.0.0.255 destination 192.168.222.0 0.0.0.255

# On the responder:

[Sysname] display acl 3000

Advanced IPv4 ACL 3000, 1 rule,

ACL's step is 5

rule 0 permit ip source 192.168.222.71 0 destination 192.168.222.5 0

3. Verify that the IPsec policy has a remote address and an IPsec transform set configured and that the IPsec transform set has all necessary settings configured.

If, for example, the IPsec policy has no remote address configured, the IPsec SA negotiation will fail:

[Sysname] display ipsec policy

-------------------------------------------

IPsec Policy: policy1

Interface: GigabitEthernet1/0/1

-------------------------------------------

-----------------------------

Sequence number: 1

Mode: ISAKMP

-----------------------------

Security data flow: 3000

Selector mode: aggregation

Local address: 192.168.222.5

Remote address:

Transform set: transform1

IKE profile: profile1

SA duration(time based):

SA duration(traffic based):

SA idle time:

Solution

1. If the IPsec policy specifies an IKE profile but no matching IKE profiles was found in IKE negotiation, perform one of the following tasks on the responder:

¡ Remove the specified IKE profile from the IPsec policy.

¡ Modify the specified IKE profile to match the IKE profile of the initiator.

2. If the flow range defined by the responder's ACL is smaller than that defined by the initiator's ACL, modify the responder's ACL so the ACL defines a flow range equal to or greater than that of the initiator's ACL.

For example:

[Sysname] display acl 3000

Advanced IPv4 ACL 3000, 2 rules,

ACL's step is 5

rule 0 permit ip source 192.168.222.0 0.0.0.255 destination 192.168.222.0 0.0.0.255

3. Configure the missing settings (for example, the remote address).

4. If the problem persists, contact H3C Support.

Configuring IKEv2

About IKEv2

Internet Key Exchange version 2 (IKEv2) is an enhanced version of IKEv1. The same as IKEv1, IKEv2 has a set of self-protection mechanisms and can be used on insecure networks for reliable identity authentication, key distribution, and IPsec SA negotiation. IKEv2 provides stronger protection against attacks and higher key exchange ability and needs fewer message exchanges than IKEv1.

IKEv2 negotiation process

Compared with IKEv1, IKEv2 simplifies the negotiation process and is much more efficient.

IKEv2 defines three types of exchanges: initial exchanges, CREATE_CHILD_SA exchange, and INFORMATIONAL exchange.

As shown in Figure 10, IKEv2 uses two exchanges during the initial exchange process: IKE_SA_INIT and IKE_AUTH, each with two messages.

· IKE_SA_INIT exchange—Negotiates IKE SA parameters and exchanges keys.

· IKE_AUTH exchange—Authenticates the identity of the peer and establishes IPsec SAs.

After the four-message initial exchanges, IKEv2 sets up one IKE SA and one pair of IPsec SAs. For IKEv1 to set up one IKE SA and one pair of IPsec SAs, it must go through two phases that use a minimum of six messages.

To set up one more pair of IPsec SAs within the IKE SA, IKEv2 goes on to perform an additional two-message exchange—the CREATE_CHILD_SA exchange. One CREATE_CHILD_SA exchange creates one pair of IPsec SAs. IKEv2 also uses the CREATE_CHILD_SA exchange to rekey IKE SAs and Child SAs.

IKEv2 uses the INFORMATIONAL exchange to convey control messages about errors and notifications.

Figure 10 IKEv2 Initial exchange process

New features in IKEv2

DH guessing

In the IKE_SA_INIT exchange, the initiator guesses the DH group that the responder is most likely to use and sends it in an IKE_SA_INIT request message. If the initiator's guess is correct, the responder responds with an IKE_SA_INIT response message and the IKE_SA_INIT exchange is finished. If the guess is wrong, the responder responds with an INVALID_KE_PAYLOAD message that contains the DH group that it wants to use. The initiator then uses the DH group selected by the responder to reinitiate the IKE_SA_INIT exchange. The DH guessing mechanism allows for more flexible DH group configuration and enables the initiator to adapt to different responders.

Cookie challenging

Messages for the IKE_SA_INIT exchange are in plain text. An IKEv1 responder cannot confirm the validity of the initiators and must maintain half-open IKE SAs, which makes the responder susceptible to DoS attacks. An attacker can send a large number of IKE_SA_INIT requests with forged source IP addresses to the responder, exhausting the responder's system resources.

IKEv2 introduces the cookie challenging mechanism to prevent such DoS attacks. When an IKEv2 responder maintains a threshold number of half-open IKE SAs, it starts the cookie challenging mechanism. The responder generates a cookie and includes it in the response sent to the initiator. If the initiator initiates a new IKE_SA_INIT request that carries the correct cookie, the responder considers the initiator valid and proceeds with the negotiation. If the carried cookie is incorrect, the responder terminates the negotiation.

The cookie challenging mechanism automatically stops working when the number of half-open IKE SAs drops below the threshold.

IKEv2 SA rekeying

For security purposes, both IKE SAs and IPsec SAs have a lifetime and must be rekeyed when the lifetime expires. An IKEv1 SA lifetime is negotiated. An IKEv2 SA lifetime, in contrast, is configured. If two peers are configured with different lifetimes, the peer with the shorter lifetime always initiates the SA rekeying. This mechanism reduces the possibility that two peers will simultaneously initiate a rekeying. Simultaneous rekeying results in redundant SAs and SA status inconsistency on the two peers.

IKEv2 message retransmission

Unlike IKEv1 messages, IKEv2 messages appear in request/response pairs. IKEv2 uses the Message ID field in the message header to identify the request/response pair. If an initiator sends a request but receives no response with the same Message ID value within a specific period of time, the initiator retransmits the request.

It is always the IKEv2 initiator that initiates the retransmission, and the retransmitted message must use the same Message ID value.

Protocols and standards

· RFC 2408, Internet Security Association and Key Management Protocol (ISAKMP)

· RFC 4306, Internet Key Exchange (IKEv2) Protocol

· RFC 4718, IKEv2 Clarifications and Implementation Guidelines

· RFC 2412, The OAKLEY Key Determination Protocol

· RFC 5996, Internet Key Exchange Protocol Version 2 (IKEv2)

IKEv2 tasks at a glance

To configure IKEv2, perform the following tasks:

1. Configuring an IKEv2 profile

a. Creating an IKEv2 profile

b. Specifying the local and remote identity authentication methods

c. Configuring the IKEv2 keychain or PKI domain

d. Configuring the local ID for the IKEv2 profile

e. Configuring peer IDs for the IKEv2 profile

f. Configuring optional features for the IKEv2 profile

2. Configuring an IKEv2 policy

3. Configuring an IKEv2 proposal

If you specify an IKEv2 proposal in an IKEv2 policy, you must configure the IKEv2 proposal.

4. Configuring an IKEv2 keychain

This task is required when either end or both ends use the preshared key authentication method.

5. (Optional.) Enabling the cookie challenging feature

The cookie challenging feature takes effect only on IKEv2 responders.

6. (Optional.) Configuring the IKEv2 DPD feature

7. (Optional.) Configuring the IKEv2 NAT keepalive feature

8. (Optional.) Configuring IKEv2 address pools

Prerequisites for IKEv2 configuration

Determine the following parameters prior to IKEv2 configuration:

· The strength of the algorithms for IKEv2 negotiation, including the encryption algorithms, integrity protection algorithms, PRF algorithms, and DH groups. Different algorithms provide different levels of protection. A stronger algorithm means better resistance to decryption of protected data but requires more resources. Typically, the longer the key, the stronger the algorithm.

· The local and remote identity authentication methods.

¡ To use the preshared key authentication method, you must determine the preshared key.

¡ To use the RSA digital signature authentication method, you must determine the PKI domain for the local end to use. For information about PKI, see "Configuring PKI."

Configuring an IKEv2 profile

Creating an IKEv2 profile

About this task

An IKEv2 profile is intended to provide a set of parameters for IKEv2 negotiation.

Procedure

1. Enter system view.

system-view

2. Create an IKEv2 profile and enter its view.

ikev2 profile profile-name

Specifying the local and remote identity authentication methods

Restrictions and guidelines

The local and remote identity authentication methods must both be specified and they can be different. You can specify only one local identity authentication method and multiple remote identity authentication methods.

Procedure

1. Enter system view.

system-view

2. Enter IKEv2 profile view.

ikev2 profile profile-name

3. Specify the local and remote identity authentication methods.

authentication-method { local | remote } { dsa-signature | ecdsa-signature | pre-share | rsa-signature }

By default, no local or remote identity authentication method is configured.

Configuring the IKEv2 keychain or PKI domain

Restrictions and guidelines

Configure the IKEv2 keychain or PKI domain for the IKEv2 profile to use. To use digital signature authentication, configure a PKI domain. To use preshared key authentication, configure an IKEv2 keychain.

Procedure

1. Enter system view.

system-view

2. Enter IKEv2 profile view.

ikev2 profile profile-name

3. Specify the keychain for preshared key authentication or the PKI domain used to request a certificate for digital signature authentication.

¡ Specify the keychain.

keychain keychain-name

¡ Specify the PKI domain.

certificate domain domain-name [ sign | verify ]

By default, no IKEv2 keychain or PKI domain is specified for an IKEv2 profile.

Configuring the local ID for the IKEv2 profile

Restrictions and guidelines

For digital signature authentication, the device can use an ID of any type. If the local ID is an IP address that is different from the IP address in the local certificate, the device uses the FQDN as the local ID. The FQDN is the device name configured by using the sysname command.

For preshared key authentication, the device can use an ID of any type other than the DN.

Procedure

1. Enter system view.

system-view

2. Enter IKEv2 profile view.

ikev2 profile profile-name

3. Configure the local ID.

identity local { address { ipv4-address | ipv6 ipv6-address } | dn | email email-string | fqdn fqdn-name | key-id key-id-string }

By default, no local ID is configured, and the device uses the IP address of the interface where the IPsec policy applies as the local ID.

Configuring peer IDs for the IKEv2 profile

About this task

Perform this task to configure the peer ID for IKEv2 profile matching. When the device needs to select an IKEv2 profile for IKEv2 negotiation with a peer, it compares the received peer ID with the peer IDs of its local IKE profiles. If a match is found, it uses the IKEv2 profile with the matching peer ID for negotiation. IKEv2 profiles will be compared in descending order of their priorities.

Procedure

1. Enter system view.

system-view

2. Enter IKEv2 profile view.

ikev2 profile profile-name

3. Configure a peer ID.

You must configure a minimum of one peer ID on each of the two peers.

Configuring optional features for the IKEv2 profile

1. Enter system view.

system-view

2. Enter IKEv2 profile view.

ikev2 profile profile-name

3. Configure optional features as needed.

¡ Configure IKEv2 DPD.

dpd interval interval [ retry seconds ] { on-demand | periodic }

By default, IKEv2 DPD is not configured for an IKEv2 profile and an IKEv2 profile uses the DPD settings configured in system view. If IKEv2 DPD is not configured in system view either, the device does not perform dead IKEv2 peer detection.

¡ Specify the local interface or IP address to which the IKEv2 profile can be applied.

match local address { interface-type interface-number | ipv4-address | ipv6 ipv6-address }

By default, an IKEv2 profile can be applied to any local interface or local IP address.

Use this command to specify which address or interface can use the IKEv2 profile for IKEv2 negotiation. Specify the local address configured in IPsec policy or IPsec policy template view (using the local-address command) for this command. If no local address is configured, specify the IP address of the interface that uses the IPsec policy.

¡ Specify a priority for the IKEv2 profile.

priority priority

By default, the priority of an IKEv2 profile is 100.

When the device needs to select an IKEv2 profile for IKEv2 negotiation with a peer, it compares the received peer ID with the peer ID of its local IKEv2 profiles in descending order of their priorities

¡ Set the IKEv2 SA lifetime for the IKEv2 profile.

sa duration seconds

By default, the IKEv2 SA lifetime is 86400 seconds.

The local and remote ends can use different IKEv2 SA lifetimes and they do not negotiate the lifetime. The end with a smaller SA lifetime will initiate an SA negotiation when the lifetime expires.

¡ Set the IKEv2 NAT keepalive interval.

nat-keepalive seconds

By default, the global IKEv2 NAT keepalive setting is used.

Configure this command when the device is behind a NAT gateway. The device sends NAT keepalive packets regularly to its peer to prevent the NAT session from being aged because of no matching traffic.

¡ Enable the configuration exchange feature.

config-exchange { request | set { accept | send } }

By default, all configuration exchange options are disabled.

This feature applies to scenarios where the headquarters and branches communicate through virtual tunnels. It enables exchanges of IP address request and set messages between the IPsec gateway at a branch and the IPsec gateway at the headquarters.

Table 2 Parameter descriptions

Parameter	Description
request	Enables the IPsec gateway at a branch to submit IP address request messages to the IPsec gateway at the headquarters.
set accept	Enables the IPsec gateway at a branch to accept the IP addresses pushed by the IPsec gateway at the headquarters.
set send	Enables the IPsec gateway at the headquarters to push IP addresses to IPsec gateways at branches.

¡ Enable AAA authorization.

aaa authorization domain domain-name username user-name

By default, AAA authorization is disabled.

The AAA authorization feature enables IKEv2 to request authorization attributes, such as the IKEv2 address pool, from AAA. IKEv2 uses the address pool to assign IP addresses to remote users. For more information about AAA authorization, see AAA in User Access and Authentication Configuration Guide.

Configuring an IKEv2 policy

About this task

During the IKE_SA_INIT exchange, each end tries to find a matching IKEv2 policy, using the IP address of the local security gateway as the matching criterion.

· If IKEv2 policies are configured, IKEv2 searches for an IKEv2 policy that uses the IP address of the local security gateway. If no IKEv2 policy uses the IP address or the policy is using an incomplete proposal, the IKE_SA_INIT exchange fails.

· If no IKEv2 policy is configured, IKEv2 uses the system default IKEv2 policy default.

The device matches IKEv2 policies in the descending order of their priorities. To determine the priority of an IKEv2 policy:

1. First, the device examines the existence of the match local address command. An IKEv2 policy with the match local address command configured has a higher priority.

2. If a tie exists, the device compares the priority numbers. An IKEv2 policy with a smaller priority number has a higher priority.

3. If a tie still exists, the device prefers an IKEv2 policy configured earlier.

Procedure

1. Enter system view.

system-view

2. Create an IKEv2 policy and enter its view.

ikev2 policy policy-name

By default, an IKEv2 policy named default exists.

3. Specify the local interface or address used for IKEv2 policy matching.

match local address { interface-type interface-number | ipv4-address | ipv6 ipv6-address }

By default, no local interface or address is used for IKEv2 policy matching, and the policy matches any local interface or address.

4. Specify an IKEv2 proposal for the IKEv2 policy.

proposal proposal-name

By default, no IKEv2 proposal is specified for an IKEv2 policy.

5. Specify a priority for the IKEv2 policy.

priority priority

By default, the priority of an IKEv2 policy is 100.

Configuring an IKEv2 proposal

About this task

An IKEv2 proposal contains security parameters used in IKE_SA_INIT exchanges, including the encryption algorithms, integrity protection algorithms, PRF algorithms, and DH groups. An algorithm specified earlier has a higher priority.

Restrictions and guidelines

A complete IKEv2 proposal must have at least one set of security parameters, including one encryption algorithm, one integrity protection algorithm, one PRF algorithm, and one DH group.

You can specify multiple IKEv2 proposals for an IKEv2 policy. A proposal specified earlier has a higher priority.

Procedure

1. Enter system view.

system-view

2. Create an IKEv2 proposal and enter its view.

ikev2 proposal proposal-name

By default, an IKEv2 proposal named default exists.

The default proposal uses the following settings:

¡ Encryption algorithms AES-CBC-128 and 3DES.

¡ Integrity protection algorithms HMAC-SHA1 and HMAC-MD5.

¡ PRF algorithms HMAC-SHA1 and HMAC-MD5.

¡ DH groups 2 and 5.

3. Specify the encryption algorithms.

encryption { 3des-cbc | aes-cbc-128 | aes-cbc-192 | aes-cbc-256 | aes-ctr-128 | aes-ctr-192 | aes-ctr-256 | camellia-cbc-128 | camellia-cbc-192 | camellia-cbc-256 | des-cbc } *

By default, an IKEv2 proposal does not have any encryption algorithms.

4. Specify the integrity protection algorithms.

integrity { aes-xcbc-mac | md5 | sha1 | sha256 | sha384 | sha512 } *

By default, an IKEv2 proposal does not have any integrity protection algorithms.

5. Specify the DH groups.

dh { group1 | group14 | group2 | group24 | group5 | group19 | group20 } *

By default, an IKEv2 proposal does not have any DH groups.

6. Specify the PRF algorithms.

prf { aes-xcbc-mac | md5 | sha1 | sha256 | sha384 | sha512 } *

By default, an IKEv2 proposal uses the integrity protection algorithms as the PRF algorithms.

Configuring an IKEv2 keychain

About this task

An IKEv2 keychain specifies the preshared keys used for IKEv2 negotiation.

An IKEv2 keychain can have multiple IKEv2 peers. Each peer has a symmetric preshared key or an asymmetric preshared key pair, and information for identifying the peer (such as the peer's host name, IP address or address range, or ID).

An IKEv2 negotiation initiator uses the peer host name or IP address/address range as the matching criterion to search for a peer. A responder uses the peer host IP address/address range or ID as the matching criterion to search for a peer.

Procedure

1. Enter system view.

system-view

2. Create an IKEv2 keychain and enter its view.

ikev2 keychain keychain-name

3. Create an IKEv2 peer and enter its view.

peer name

4. Configure a host name for the peer:

hostname name

By default, no host name is configured for an IKEv2 peer.

5. Configure a host IP address or address range for the peer:

address { ipv4-address [ mask | mask-length ] | ipv6 ipv6-address [ prefix-length ] }

By default, no host IP address or address range is configured for an IKEv2 peer.

You must configure different host IP addresses/address ranges for different peers.

6. Configure an ID for the peer:

identity { address { ipv4-address | ipv6 { ipv6-address } } | fqdn fqdn-name | email email-string | key-id key-id-string }

By default, no identity information is configured for an IKEv2 peer.

7. Configure a preshared key for the peer.

pre-shared-key [ local | remote ] { ciphertext | plaintext } string

By default, an IKEv2 peer does not have a preshared key.

Configure global IKEv2 parameters

Enabling the cookie challenging feature

About this task

Enable cookie challenging on responders to protect them against DoS attacks that use a large number of source IP addresses to forge IKE_SA_INIT requests.

Procedure

1. Enter system view.

system-view

2. Enable IKEv2 cookie challenging.

ikev2 cookie-challenge number

By default, IKEv2 cookie challenging is disabled.

Configuring the IKEv2 DPD feature

About this task

IKEv2 DPD detects dead IKEv2 peers in periodic or on-demand mode.

· Periodic IKEv2 DPD—Verifies the liveness of an IKEv2 peer by sending DPD messages at regular intervals.

· On-demand IKEv2 DPD—Verifies the liveness of an IKEv2 peer by sending DPD messages before sending data.

¡ Before the device sends data, it identifies the time interval for which the last IPsec packet has been received from the peer. If the time interval exceeds the DPD interval, it sends a DPD message to the peer to detect its liveliness.

¡ If the device has no data to send, it never sends DPD messages.

Restrictions and guidelines

If you configure IKEv2 DPD in both IKEv2 profile view and system view, the IKEv2 DPD settings in IKEv2 profile view apply. If you do not configure IKEv2 DPD in IKEv2 profile view, the IKEv2 DPD settings in system view apply.

Procedure

1. Enter system view.

system-view

2. Configure global IKEv2 DPD.

ikev2 dpd interval interval [ retry seconds ] { on-demand | periodic }

By default, global DPD is disabled.

Configuring the IKEv2 NAT keepalive feature

About this task

Configure this feature on the IKEv2 gateway behind the NAT device. The gateway then sends NAT keepalive packets regularly to its peer to keep the NAT session alive, so that the peer can access the device.

The NAT keepalive interval must be shorter than the NAT session lifetime.

This feature takes effect after the device detects the NAT device.

Procedure

1. Enter system view.

system-view

2. Set the IKEv2 NAT keepalive interval.

ikev2 nat-keepalive seconds

By default, the IKEv2 NAT keepalive interval is 10 seconds.

Configuring IKEv2 address pools

About this task

To perform centralized management on remote users, an IPsec gateway can use an address pool to assign private IP addresses to remote users.

You must use an IKEv2 address pool together with AAA authorization by specifying the IKEv2 address pool as an AAA authorization attribute. For more information about AAA authorization, see AAA in User Access and Authentication Configuration Guide.

Procedure

1. Enter system view.

system-view

2. Configure an IKEv2 IPv4 address pool.

ikev2 address-group group-name start-ipv4-address end-ipv4-address [ mask | mask-length ]

3. Configure an IKEv2 IPv6 address pool.

ikev2 ipv6-address-group group-name prefix prefix/prefix-len assign-len assign-len

Display and maintenance commands for IKEv2

Execute display commands in any view and reset commands in user view.

Task	Command
Display the IKEv2 policy configuration.	display ikev2 policy [ policy-name \| default ]
Display the IKEv2 profile configuration.	display ikev2 profile [ profile-name ]
Display the IKEv2 proposal configuration.	display ikev2 proposal [ name \| default ]
Display the IKEv2 SA information.	display ikev2 sa [ count \| [ { local \| remote } { ipv4-address \| ipv6 ipv6-address } ] [ verbose [ tunnel tunnel-id ] ] ]
Display IKEv2 statistics.	display ikev2 statistics
Delete IKEv2 SAs and the child SAs negotiated through the IKEv2 SAs.	reset ikev2 sa [ [ { local \| remote } { ipv4-address \| ipv6 ipv6-address } ] \| tunnel tunnel-id ] [ fast ]
Clear IKEv2 statistics.	reset ikev2 statistics

Troubleshooting IKEv2

IKEv2 negotiation failed because no matching IKEv2 proposals were found

Symptom

The IKEv2 SA is in IN-NEGO status.

<Sysname> display ikev2 sa

Tunnel ID Local Remote Status

---------------------------------------------------------------------------

5 123.234.234.124/500 123.234.234.123/500 IN-NEGO

Status:

IN-NEGO: Negotiating, EST: Established, DEL:Deleting

Analysis

Certain IKEv2 proposal settings are incorrect.

Solution

1. Examine the IKEv2 proposal configuration to see whether the two ends have matching IKEv2 proposals.

2. Modify the IKEv2 proposal configuration to make sure the two ends have matching IKEv2 proposals.

3. If the problem persists, contact H3C Support.

IPsec SA negotiation failed because no matching IPsec transform sets were found

Symptom

The display ikev2 sa command shows that the IKEv2 SA negotiation succeeded and the IKEv2 SA is in EST status. The display ipsec sa command shows that the expected IPsec SAs have not been negotiated yet.

Analysis

Certain IPsec policy settings are incorrect.

Solution

1. Examine the IPsec configuration to see whether the two ends have matching IPsec transform sets.

2. Modify the IPsec configuration to make sure the two ends have matching IPsec transform sets.

3. If the problem persists, contact H3C Support.

IPsec tunnel establishment failed

Symptom

The ACLs and IKEv2 proposals are correctly configured on both ends. The two ends cannot establish an IPsec tunnel or cannot communicate through the established IPsec tunnel.

Analysis

The IKEv2 SA or IPsec SAs on either end are lost. The reason might be that the network is unstable and the device reboots.

Solution

1. Use the display ikev2 sa command to examine whether an IKEv2 SA exists on both ends. If the IKEv2 SA on one end is lost, delete the IKEv2 SA on the other end by using the reset ikev2 sa command and trigger new negotiation. If an IKEv2 SA exists on both ends, go to the next step.

2. Use the display ipsec sa command to examine whether IPsec SAs exist on both ends. If the IPsec SAs on one end are lost, delete the IPsec SAs on the other end by using the reset ipsec sa command and trigger new negotiation.

3. If the problem persists, contact H3C Support.

14-Security Configuration Guide

Configuring IPsec

Security protocols

Transport mode

Tunnel mode

About this task

SA setup

SA aging

Authentication algorithms

Encryption algorithms

Crypto engine

IPsec-protected traffic

Configuring an ACL

Keywords in ACL rules

Mirror image ACLs

Configuring an IPsec transform set

About this task

Restrictions and guidelines

Procedure

Configuring a manual IPsec policy

Procedure

Configuring an IKE-based IPsec policy

About this task

Restrictions and guidelines for IKE-based IPsec policy configuration

Directly configuring an IKE-based IPsec policy

Configuring an IKE-based IPsec policy by using an IPsec policy template

Applying an IPsec policy to an interface

Restrictions and guidelines

Procedure

About this task

Procedure

Configuring IPsec anti-replay

About this task

Restrictions and guidelines

Procedure

About this task

Restrictions and guidelines

Procedure

Enabling QoS pre-classify

About this task

Restrictions and guidelines

Procedure

About this task

Restrictions and guidelines for DF bit configuration for IPsec packets

Configuring the DF bit of IPsec packets on an interface

Configuring the DF bit of IPsec packets globally

Restrictions and guidelines

Procedure

About this task

Procedure

About this task

Restrictions and guidelines

Procedure

Restrictions and guidelines

Procedure

About this task

Procedure

About this task

Procedure

About this task

Procedure

Display and maintenance commands for IPsec

IKE negotiation process

IKE exchange process in main mode

IKE exchange process in aggressive mode

Identity authentication

DH algorithm

PFS

Protocols and standards

IKE tasks at a glance

About this task

Procedure

About this task

Restrictions and guidelines

Procedure

Restrictions and guidelines

Procedure

Restrictions and guidelines

Procedure

Restrictions and guidelines