Security best practices

This chapter describes some techniques and best practices that you can use to improve FortiOS security.

Install the FortiGate unit in a physically secure location

A good place to start with is physical security. Install your FortiGate in a secure location, such as a locked room or one with restricted access. A restricted location prevents unauthorized users from getting physical access to the device.

If unauthorized users have physical access, they can disrupt your entire network by disconnecting your FortiGate (either by accident or on purpose). They could also connect a console cable and attempt to log into the CLI. Also, when a FortiGate unit reboots, a person with physical access can interrupt the boot process and install different firmware.

Register your product with Fortinet Support

You need to register your Fortinet product with Fortinet Support to receive customer services, such as firmware updates and customer support. You must also register your product for FortiGuard services, such as up-to-date antivirus and IPS signatures. To register your product the Fortinet Support website.

Keep your FortiOS firmware up to date

Always keep FortiOS up to date. The most recent version is the most stable and has the most bugs fixed and vulnerabilities removed. Fortinet periodically updates the FortiGate firmware to include new features and resolve important issues.

After you register your FortiGate, you can receive notifications on FortiGate GUI about firmware updates. You can update the firmware directly from the GUI or by downloading firmware updates from the Fortinet Support website.

Before you install any new firmware, be sure to follow these steps:

• Review the release notes for the latest firmware release.
• Review the Supported Upgrade Paths guide to determine the best path to take from your current version of FortiOS to the latest version.
• Back up the current configuration.

Only FortiGate administrators who have read and write privileges can upgrade the FortiOS firmware.

System administrator best practices

This section describes a collection of changes you can implement to make administrative access to the GUI and CLI more secure.

Disable administrative access to the external (Internet-facing) interface

When possible, don’t allow administration access on the external (Internet-facing) interface.

To disable administrative access, go to Network > Interfaces, edit the external interface and disable HTTPS, PING, HTTP, SSH, and TELNET under Administrative Access.

From the CLI:

config system interface edit <external-interface-name> unset allowaccess

end

Allow only HTTPS access to the GUI and SSH access to the CLI

For greater security never allow HTTP or Telnet administrative access to a FortiGate interface, only allow HTTPS and SSH access. You can change these settings for individual interfaces by going to Network > Interfaces and adjusting the administrative access to each interface.

From the CLI:

config system interface edit <interface-name> set allowaccess https ssh

end

Require TLS 1.2 for HTTPS administrator access

Use the following command to require TLS 1.2 for HTTPS administrator access to the GUI:

config system global set admin-https-ssl-versions tlsv1-2

end

TLS 1.2 is currently the most secure SSL/TLS supported version for SSL-encrypted administrator access.

Re-direct HTTP GUI logins to HTTPS

Go to System > Settings > Administrator Settings and enable Redirect to HTTPS to make sure that all attempted HTTP login connections are redirected to HTTPS.

From the CLI:

config system global set admin-https-redirect enable end

Change the HTTPS and SSH admin access ports to non-standard ports

Go to System > Settings > Administrator Settings and change the HTTPS and SSH ports.

You can change the default port configurations for HTTPS and SSH administrative access for added security. To connect to a non-standard port, the new port number must be included in the collection request. For example:

l If you change the HTTPS port to 7734, you would browse to https://<ip-address>:7734. l If you change the SSH port to 2345, you would connect to ssh admin@<ip-address>:2345 To change the HTTPS and SSH login ports from the CLI:

config system global set admin-sport 7734 set admin-ssh-port 2345

end

If you change to the HTTPS or SSH port numbers, make sure your changes do not conflict with ports used for other services.

Maintain short login timeouts

Set the idle timeout to a short time to avoid the possibility of an administrator walking away from their management computer and leaving it exposed to unauthorized personnel.

To set the administrator idle timeout, go to System > Settings and enter the amount of time for the Idle timeout. A best practice is to keep the default time of 5 minutes.

To set the administrator idle timeout from the CLI:

config system global set admintimeout 5

end

You can use the following command to adjust the grace time permitted between making an SSH connection and authenticating. The range can be between 10 and 3600 seconds, the default is 120 seconds (minutes). By shortening this time, you can decrease the chances of someone attempting a brute force attack a from being successful. For example, you could set the time to 30 seconds.

config system global set admin-ssh-grace-time 30

end

Restrict logins from trusted hosts

Setting up trusted hosts for an administrator limits the addresses from where they can log into FortiOS. The trusted hosts configuration applies to most forms of administrative access including HTTPS, SSH, and SNMP. When you identify a trusted host for an administrator account, FortiOS accepts that administrator’s login only from one of the trusted hosts. A login, even with proper credentials, from a non-trusted host is dropped.

System administrator best practices

To identify trusted hosts, go to System > Administrators, edit the administrator account, enable Restrict login to trusted hosts, and add up to ten trusted host IP addresses.

To add two trusted hosts from the CLI:

config system admin edit <administrator-name> set trustedhost1 172.25.176.23 255.255.255.255 set trustedhost2 172.25.177.0 255.255.255.0

end

Trusted host IP addresses can identify individual hosts or subnets. Just like firewall policies, FortiOS searches through the list of trusted hosts in order and acts on the first match it finds. When you configure trusted hosts, start by adding specific addresses at the top of the list. Follow with more general IP addresses. You don’t have to add addresses to all of the trusted hosts as long as all specific addresses are above all of the 0.0.0.0 0.0.0.0 addresses.

Set up two-factor authentication for administrators

FortiOS supports FortiToken and FortiToken Mobile 2-factor authentication. FortiToken Mobile is available for iOS and Android devices from their respective application stores.

Every registered FortiGate unit includes two trial tokens for free. You can purchase additional tokens from your reseller or from Fortinet.

To assign a token to an administrator, go to System > Administrators and select Enable Two-factor Authentication for each administrator.

Create multiple administrator accounts

Rather than allowing all administrators to access ForiOS with the same administrator account, you can create accounts for each person or each role that requires administrative access. This configuration allows you to track the activities of each administrator or administrative role.

If you want administrators to have different functions you can add different administrator profiles. Go to System > Admin Profiles and select Create New.

Modify administrator account lockout duration and threshold values

By default, the FortiGate sets the number of password retries at three, allowing the administrator a maximum of three attempts to log into their account before locking the account for a set amount of time.

Both the number of attempts (admin-lockout-threshold) and the wait time before the administrator can try to enter a password again (admin-lockout-duration) can be configured within the CLI.

To configure the lockout options:

config system global set admin-lockout-threshold <failed_attempts> set admin-lockout-duration <seconds>

end

The default value of admin-lockout-threshold is 3 and the range of values is between 1 and 10. The admin-lockout-duration is set to 60 seconds by default and the range of values is between 1 and 4294967295 seconds.

Global commands for stronger and more secure encryption

Keep in mind that the higher the lockout threshold, the higher the risk that someone may be able to break into the FortiGate.

Example:

To set the admin-lockout-threshold to one attempt and the admin-lockout-duration to a five minute duration before the administrator can try to log in again, enter the commands:

config system global set admin-lockout-threshold 1 set admin-lockout-duration 300 end

If the time span between the first failed login attempt and the admin-lockoutthreshold failed login attempt is less than admin-lockout-duration, the lockout will be triggered.

Rename the admin administrator account

You can improve security by renaming the admin account. To do this, create a new administrator account with the super_admin admin profile and log in as that administrator. Then go to System > Administrators and edit the admin administrator and change the User Name. Renaming the admin account makes it more difficult for an attacker to log into FortiOS.

Add administrator disclaimers

FortiOS can display a disclaimer before or after logging into the GUI or CLI (or both). In either case the administrator must read and accept the disclaimer before they can proceed.

Use the following command to display a disclaimer before logging in:

config system global set pre-login-banner enable

end

Use the following command to display a disclaimer after logging in:

config system global set post-login-banner enable

end

You can customize the replacement messages for these disclaimers by going to System > Replacement Messages. Select Extended View to view and edit the Administrator replacement messages.

From the CLI:

config system replacemsg admin pre_admin-disclaimer-text config system replacemsg admin post_admin-disclaimer-text

Global commands for stronger and more secure encryption

This section describes some best practices for employing stronger and more secure encryption.

Disable sending malware statistics to FortiGuard

Turn on global strong encryption

Enter the following command to configure FortiOS to use only strong encryption and allow only strong ciphers (AES, 3DES) and digest (SHA1) for HTTPS, SSH, TLS, and SSL functions.

config sys global set strong-crypto enable

end

Disable MD5 and CBC for SSH

In some cases, you may not be able to enable strong encryption. For example, your FortiGate may be communicating with a system that does not support strong encryption. With strong-crypto disabled you can use the following options to prevent SSH sessions with the FortiGate from using less secure MD5 and CBC algorithms:

config sys global set ssh-hmac-md5 disable set ssh-cbc-cipher disable

end

Disable static keys for TLS

You can use the following command to prevent TLS sessions from using static keys (AES128-SHA, AES256-SHA, AES128-SHA256, AES256-SHA256):

config sys global set ssl-static-key-ciphers disable

end

Require larger values for Diffie-Hellman exchanges

Larger Diffie-Hellman values result in stronger encryption. Use the following command to force Diffie-Hellman exchanges to use 8192 bit values (the highest configurable DH value).

config sys global set dh-params 8192

end

Disable sending malware statistics to FortiGuard

By default FortiOS periodically sends encrypted malware statistics to FortiGuard. The malware statistics record Antivirus, IPS, or Application Control events. This data is used to improved FortiGuard services. The malware statistics that FortiOS sends do not include any personal or sensitive customer data. The information is not shared with any external parties and is used in accordance with Fortinet’s Privacy Policy.

To disable sending malware statistics to FortiGuard, enter the following command: config system global set fds-statistics disable

end

Disable sending Security Rating statistics to FortiGuard

Security Rating is a Fortinet Security Fabric feature that allows customers to audit their Security Fabric and find and fix security problems. As part of the feature, FortiOS sends your security rating to FortiGuard every time a security rating test runs.

You can opt out of submitting Security Rating scores to FortiGuard. If you opt out you won’t be able to see how your organization’s scores compare with the scores of other organizations. Instead, an absolute score is shown. Use the following command to disable FortiGuard Security Rating result submission:

config system global set fortiguard-audit-result-submission disable

end

Disable auto USB installation

If USB installation is enabled, an attacker with physical access to a FortiGate could load a new configuration or firmware on the FortiGate using the USB port. You can disable USB installation by entering the following from the CLI:

config system auto-install set auto-install-config disable set auto-install-image disable

end

Set system time by synchronizing with an NTP server

For accurate time, use an NTP server to set system time. Synchronized time facilitates auditing and consistency between expiry dates used in expiration of certificates and security protocols.

From the GUI go to System > Settings > System Time and select Synchronize with NTP Server. By default, this causes FortiOS to synchronize with Fortinet’s FortiGuard secure NTP server.

From the CLI you can use one or more different NTP servers:

config system ntp set type custom set ntpsync enable config ntpserver edit 1 set server <ntp-server-ip>

next edit 2 set server <other-ntp-server-ip> end

Disable the maintainer admin account

Administrators with physical access to a FortiGate appliance can use a console cable and a special administrator account called maintainer to log into the CLI without a password. This feature allows you to log into a FortiGate if you have lost all administrator passwords. See Resetting a lost Admin password on the Fortinet Cookbook for details.

The maintainer account can be disabled using the following command:

config system global set admin-maintainer disable

end

Enable password policies

Go to System > Settings > Password Policy, to create a password policy that all administrators must follow. Using the available options you can define the required length of the password, what it must contain (numbers, upper and lower case, and so on) and an expiry time.

Use the password policy feature to make sure all administrators use secure passwords that meet your organization’s requirements.

Configure auditing and logging

For optimum security go to Log & Report > Log Settings enable Event Logging. For best results send log messages to FortiAnalyzer or FortiCloud.

From FortiAnalyzer or FortiCloud, you can view reports or system event log messages to look for system events that may indicate potential problems. You can also view system events by going to FortiView > System Events.

Establish an auditing schedule to routinely inspect logs for signs of intrusion and probing.

Encrypt logs sent to FortiAnalyzer/FortiManager

To keep information in log messages sent to FortiAnalyzer private, go to Log & Report > Log Settings and when you configure Remote Logging to FortiAnalyzer/FortiManager select Encrypt log transmission.

From the CLI.

config log {fortianalyzer | fortianalyzer2 | fortianalyzer3} setting set enc-algorithm high end

Disable unused interfaces

To disable an interface from the GUI, go to Network > Interfaces. Edit the interface to be disabled and set Interface State to Disabled.

From the CLI, to disable the port21 interface:

config system interface edit port21 set status down

end

Disable unused protocols on interfaces

You can use the config system interface command to disable unused protocols that attackers may attempt to use to gather information about a FortiGate unit. Many of these protocols are disabled by default. Using the config system interface command you can see the current configuration of each of these options for the selected interface and then choose to disable them if required.

config system interface edit <interface-name> set dhcp-relay-service disable set pptp-client disable set arpforward disable set broadcast-forward disable set l2forward disable set icmp-redirect disable set vlanforward disable set stpforward disable set ident-accept disable set ipmac disable set netbios-forward disable set security-mode none set device-identification disable set lldp-transmission disable end

Use local-in policies to close open ports or restrict access

You can also use local-in policies to close open ports or otherwise restrict access to FortiOS.

Close ICMP ports

Use the following command to close all ICMP ports on the WAN1 interface. The following example blocks traffic that matches the ICMP_ANY firewall service.

config firewall local-in-policy edit 1 set intf wan1 set scraddr all set dstaddr all set action deny set service ICMP_ANY set schedule always

end

Close the BGP port

Use the following command to close the BGP port on the wan1 interface. The following example blocks traffic that matches the BGP firewall service.

config firewall local-in-policy edit 1 set intf wan1 set scraddr all set dstaddr all

Use local-in policies to close open ports or restrict access

set action deny set service BGP set schedule always end

IP, TCP, and UDP load balancing

You can load balance all IP, TCP or UDP sessions accepted by the security policy that includes a load balancing virtual server with the type set to IP, TCP, or UDP. Traffic with destination IP and port that matches the virtual server IP and port is load balanced. For these protocol-level load balancing virtual servers you can select a load balance method and add real servers and health checking. However, you can’t configure persistence, HTTP multiplexing and SSL offloading.

Example HTTP load balancing to three real web servers

In this example, a virtual web server with IP address 192.168.37.4 on the Internet, is mapped to three real web servers connected to the FortiGate unit dmz1 interface. The real servers have IP addresses 10.10.123.42, 10.10.123.43, and 10.10.123.44. The virtual server uses the First Alive load balancing method. The configuration also includes an HTTP health check monitor that includes a URL used by the FortiGate unit for get requests to monitor the health of the real servers.

Connections to the virtual web server at IP address 192.168.37.4 from the Internet are translated and load balanced to the real servers by the FortiGate unit. First alive load balancing directs all sessions to the first real server. The computers on the Internet are unaware of this translation and load balancing and see a single virtual server at IP address 192.168.37.4 rather than the three real servers behind the FortiGate unit.

Virtual server configuration example

GUI configuration

Use the following procedures to configure this load balancing setup from the GUI.

To add an HTTP health check monitor

In this example, the HTTP health check monitor includes the URL “/index.html” and the Matched Phrase “Fortinet products”.

1. Go to Policy & Objects > Health Check.
2. Select Create New.
3. Add an HTTP health check monitor that sends get requests to http://<real_server_IP_address>/index.html and searches the returned web page for the phrase “Fortinet products”.

4. Select OK.

To add the HTTP virtual server and the real servers

1. Go to Policy & Objects > Virtual Servers.
2. Select Create New.
3. Add an HTTP virtual server that allows users on the Internet to connect to the real servers on the internal network.

In this example, the FortiGate wan1 interface is connected to the Internet.

4. Add three real servers to the virtual server. Each real server must include the IP address of a real server on the internal network.

Configuration for the first real server.

Configuration for the second real server.

Configuration for the third real server.

HTTP load balancing to three real web servers

To add the virtual server to a security policy

Add a wan1 to dmz1 security policy that uses the virtual server so that when users on the Internet attempt to connect to the web server’s IP address, packets pass through the FortiGate unit from the wan1 interface to the dmz1 interface. The virtual IP translates the destination address of these packets from the virtual server IP address to the real server IP addresses.

1. Go to Policy & Objects > IPv4 Policy.
2. Select Create New.
3. Configure the security policy:

4. Select other security policy options as required.
5. Select OK.

SSL/TLS load balancing

In a firewall load balancing virtual server configuration, you can select SSL to load balance only SSL and TLS sessions. The virtual server will load balance SSL and TLS sessions received at the virtual server interface with destination IP address that matches the configured virtual server IP and destination port number that matches the configured virtual server port. Change this port to match the destination port of the sessions to be load balanced.

For SSL load balancing you can also set persistence to SSL session ID. Persistence is achieved by the FortiGate unit sending all sessions with the same SSL session ID to the same real server. When you configure persistence, the FortiGate unit load balances a new session to a real server according to the Load Balance Method. If the session has an SSL session ID, the FortiGate unit sends all subsequent sessions with the same SSL session ID to the same real server.

SSL/TLS offloading

Use SSL offloading to accelerate clients’ SSL or HTTPS connections to real servers by using the FortiGate unit to perform SSL/TLS operations (offloading them from the real servers using the FortiGate unit’s SSL acceleration hardware). FortiGate units can offload most versions of SSL/TLS, including SSL 3.0, TLS 1.0 and TLS 1.2. SSL/TLS offloading is available on FortiGate units that support SSL acceleration.

To configure SSL offloading from the GUI go to Policy & Objects > Virtual Servers. Add a virtual server and set the type to HTTPS or SSL and select the SSL offloading type (Client <-> FortiGate or Full).

Select Client <-> FortiGate to apply hardware accelerated SSL/TLS processing only to the part of the connection between the client and the FortiGate unit. This mode is called half mode SSL offloading. The segment between the FortiGate unit and the server will use clear text communications. This results in best performance, but cannot be used in failover configurations where the failover path does not have an SSL accelerator.

Select Full to apply hardware accelerated SSL processing to both parts of the connection: the segment between client and the FortiGate unit, and the segment between the FortiGate unit and the server. The segment between the FortiGate unit and the server uses encrypted communications, but the handshakes are abbreviated. This is not as efficient as half mode SSL offloading, but still improves performance. As well, full-mode SSL offloading can be used in failover configurations where the failover path does not have an SSL accelerator. If the server is already configured to use SSL, this also enables SSL acceleration without requiring changes to the server’s configuration.

SSL Offloading modes (Half Mode and Full Mode)

Configuring SSL offloading also requires selecting a certificate to use for the SSL offloading sessions. SSL offloading supports key sizes up to 4096. FortiGate models with CP9 processors support 3072 and 4096 DH bit sizes in hardware. All FortiGate models up to and including those with CP8 processors only support offloading DH bit sizes up to 2048 so any sizes larger than that are done in software and thus are relatively resource intensive

The following CLI command shows an example half mode HTTPS SSL offloading configuration. In the example the ssl-mode option sets the SSL offload mode to half (which is the default mode).

config firewall vip edit Vserver-ssl-offload set type server-load-balance set server-type https set ldb-method round-robin set extip 172.20.120.30 set extintf wan1 set extport 443

Separate virtual-server client and server TLS version and cipher configuration

set persistence ssl-session-id set ssl-mode half set ssl-certificate my-cert set monitor tcp-mon-1 config realservers edit 1 set ip 10.31.101.30 set port 443

next edit 2 set ip 10.31.101.40 set port 443

end

end

Separate virtual-server client and server TLS version and cipher configuration

In some cases, you may want the to use different versions of SSL or TLS on the client to FortiGate connection than on the FortiGate to server connection. For example, you may want to use the FortiGate to protect a legacy SSL 3.0 or TLS 1.0 server while making sure that client to FortiGate connections must always use the higher level of protection offered by TLS 1.1 or greater. Also, in some cases you might want to protect a server that only has weak ciphers (for example, DES or RC4) while making sure that all connections between the FortiGate and the client use a strong cipher for better protection.

The following options are available when configuring server load balancing for HTTPS sessions configured with the following command:

config firewall vip edit server-name set type server-load-balance set server-type https set ssl-mode full …

Setting the SSL/TLS versions to use for server and client connections

The ssl-server-min-version, ssl-server-max-version, ssl-min-version and ssl-maxversion configuration options allow the minimum and maximum SSL/TLS versions for the client to FortiGate connection to be independent of the FortiGate to server configuration. By default these options are both set to client and the configured ssl-min-version and ssl-max-version settings are applied to both the client and the server connection.

You can change the ssl-server-min-version and ssl-server-max-version to apply different options to the server connection. The ssl-min-version and ssl-max-version settings are still applied to the client connection. If you set the ssl-server-min-version and ssl-server-max-version to an explicit version then both must be set to an explicit version.

The ssl-server-min-version and ssl-server-max-version options allow you to specify the minimum and maximum SSL/TLS versions the FortiGate will offer to the server (in the record header of the ClientHello) when performing full mode SSL offloading and thus the minimum and maximum SSL/TLS versions the FortiGate accepts from the server (in a ServerHello). If the server responds with a version in its ServerHello Setting the SSL/TLS cipher choices for server and client connections

that is lower than ssl-server-min-version or higher than the ssl-server-max-version then the FortiGate terminates the connection.

Command syntax is:

config firewall vip edit server-name set type server-load-balance set server-type https set ssl-mode full set ssl-min-version {ssl-3.0 | tls-1.0 | tls-1.1 | tls-1.2} set ssl-max-version {ssl-3.0 | tls-1.0 | tls-1.1 | tls-1.2} set ssl-server-min-version {ssl-3.0 | tls-1.0 | tls-1.1 | tls-1.2 | client} set ssl-server-max-version {ssl-3.0 | tls-1.0 | tls-1.1 | tls-1.2 | client}

Setting the SSL/TLS cipher choices for server and client connections

The ssl-algorithm and ssl-server-algorithm configuration options allow the cipher choice for the FortiGate to server connection to be independent of the client to FortiGate connection. By default, sslserver-algorithm is set to client and the configured ssl-algorithm setting is applied to both the client and the server connection.

You can change the ssl-server-algorithm to apply different options to the server connection. The sslalgorithm setting is still applied to the client connection. The following ssl-server-algorithm options are available:

• high, offer AES or 3DES cypher suites in the ServerHello l medium, use AES, 3DES, or RC4 cypher suites in the ServerHello l low, use AES, 3DES, RC4, or DES cypher suites in the ServerHello l custom, specifiy custom cypher suites using the config ssl-server-cipher-suites and offer these custom cypher suites in the ServerHello.
• client, offer the cypher suites in the ServerHello that are offered in the ClientHello.

Command syntax is:

config firewall vip edit server-name set type server-load-balance set server-type https set ssl-mode full

set ssl-algorithm {high | medium | low | custom}

set ssl-server-algorithm {high | medium | low | custom | client}

If you set ssl-server-algorithm to custom, the syntax is: config firewall vip edit server-name set type server-load-balance set server-type https set ssl-mode full

set ssl-server-algorithm custom config ssl-server-cipher-suites edit 10 set cipher <cipher-suite>

set versions {ssl-3.0 | tls-1.0 | tls-1.1 | tls-1.2}

Protection from TLS protocol downgrade attacks

next edit 20 set cipher <cipher-suite>

set versions {ssl-3.0 | tls-1.0 | tls-1.1 | tls-1.2} end

Protection from TLS protocol downgrade attacks

The ssl-client-fallback option, when enabled (the default configuration), performs downgrade attack prevention (RFC 7507).

Command syntax is:

config firewall vip edit server-name set type server-load-balance set server-type https set ssl-client-fallback {disable | enable}

Setting 3072- and 4096-bit Diffie-Hellman values

The ssl-dh-bits option allows you to specify the number of bits of the prime number used in the DiffieHellman exchange for RSA encryption of the SSL connection. Larger prime numbers are associated with greater cryptographic strength. You can set DH values from 768 to 4096 bits.

Command syntax is:

config firewall vip edit server-name set type server-load-balance set server-type https set ssl-dh-bits {768 | 1024 | 1536 | 2048 | 3072 | 4096}

Setting the DH bits to 2048 only provides the equivalent of a symmetric cipher in the range of 112 – 128 bits. This means that if AES 256 is used then the weakest point is the DH of 2048 and a value of at least 3072 should be use if the goal is to have 256 bits of security.

FortiGate models with CP9 processors support 3072 and 4096 DH bit sizes in hardware. All FortiGate models up to and including those with CP8 processors only support offloading DH bit sizes up to 2048 so any sizes larger than that are done in software and thus are relatively resource intensive.

Additional SSL load balancing and SSL offloading options

The following SSL load balancing and SSL offloading options are only available from the CLI:

ssl-client-session-state-max <sessionstates_int>

Enter the maximum number of SSL session states to keep for the segment of the SSL connection between the client and the FortiGate unit.

ssl-client-session-state-timeout <timeout_int>

Additional SSL load balancing and SSL offloading options

Enter the number of minutes to keep the SSL session states for the segment of the SSL connection between the client and the FortiGate unit.

ssl-client-session-state-type {both | client | disable | time}

Select which method the FortiGate unit should use when deciding to expire SSL sessions for the segment of the SSL connection between the client and the FortiGate unit.

• both: Select to expire SSL session states when either ssl-client-session-state-max or ssl-clientsession-state-timeout is exceeded, regardless of which occurs first. l count: Select to expire SSL session states when ssl-client-session-state-max is exceeded.
• disable: Select to keep no SSL session states.
• time: Select to expire SSL session states when ssl-client-session-state-timeout is exceeded.

ssl-http-location-conversion {enable | disable}

Select to replace http with https in the reply’s Location HTTP header field. For example, in the reply,

Location: http://example.com/ would be converted to Location: https://example.com/ ssl-http-match-host {enable | disable}

Enable (the default) to apply Location conversion to the reply’s HTTP header only if the host name portion of Location matches the request’s Host field, or, if the Host field does not exist, the host name portion of the request’s URI.

If disabled, conversion occurs regardless of whether the host names in the request and the reply match.

For example, if host matching is enabled, and a request contains Host: example.com and the reply contains Location: http://example.cc/, the Location field does not match the host of the original request and the reply’s Location field remains unchanged. If the reply contains Location: http://example.com/, however, then the FortiGate unit detects the matching host name and converts the reply field to Location: https://example.com/.

This option appears only if ssl-http-location-conversion is enable.

ssl-send-empty-frags {enable | disable}

Select to precede the record with empty fragments to protect from attacks on CBC IV. You might disable this option if SSL acceleration will be used with an old or buggy SSL implementation which cannot properly handle empty fragments.

ssl-server-session-state-max <sessionstates_int>

Enter the maximum number of SSL session states to keep for the segment of the SSL connection between the server and the FortiGate unit.

ssl-server-session-state-timeout <timeout_int>

Enter the number of minutes to keep the SSL session states for the segment of the SSL connection between the server and the FortiGate unit. This option appears only if ssl-mode is full.

ssl-server-session-state-type {both | count | disable | time}

Select which method the FortiGate unit should use when deciding to expire SSL sessions for the segment of the SSL connection between the server and the FortiGate unit. This option appears only if ssl-mode is full.

• both: Select to expire SSL session states when either ssl-server-session-state-max or ssl-serversession-state-timeout is exceeded, regardless of which occurs first. l count: Select to expire SSL session states when ssl-server-session-state-max is exceeded.
• disable: Select to keep no SSL session states. l time: Select to expire SSL session states when ssl-server-session-state-timeout is exceeded.

SSL offloading support for Internet Explorer 6

In some cases the Internet Explorer 6 web browser may be able to access real servers. To resolve this issue, disable the ssl-send-empty-frags option:

config firewall vip edit vip_name set type server-load-balance set server-type https set ssl-send-empty-frags disable

end

You can disable this option if SSL acceleration will be used with an old or buggy SSL implementation that cannot properly handle empty fragments.

Selecting the cipher suites available for SSL load balancing

You can use the following command to view the complete list of cipher suites available for SSL offloading:

config firewall vip edit <vip-name> set type server-load-balance set server-type https set ssl-algorithm custom config ssl-cipher-suites edit 0 set cipher ?

In most configurations the matching cipher suite is automatically selected but you can limit the set of cipher suites that are available for a given SSL offloading configuration. For example, use the following command to limit an SSL load balancing configuration to use the three cipher suites that support ChaCha20 and Poly1305:

config firewall vip edit <vip-name> set type server-load-balance set server-type https set ssl-algorithm custom config ssl-cipher-suites edit 1

set cipher TLS-ECDHE-RSA-WITH-CHACHA20-POLY1305-SHA256 next edit 2

set cipher TLS-ECDHE-ECDSA-WITH-CHACHA20-POLY1305-SHA256 next edit 3

set cipher TLS-DHE-RSA-WITH-CHACHA20-POLY1305-SHA256 end end

Disabling SSL/TLS re-negotiation

The vulnerability CVE-2009-3555 affects all SSL/TLS servers that support re-negotiation. FortiOS when configured for SSL/TLS offloading is operating as a SSL/TLS server. The IETF is working on a TLS protocol change that will fix the problem identified by CVE-2009-3555 while still supporting re-negotiation. Until that protocol change is available, you can use the ssl-client-renegotiation option to disable support for SSL/TLS re-negotiation. The default value of this option is allow, which allows an SSL client to renegotiate. You can change the setting to deny to abort any attempts by an SSL client to renegotiate. If you select deny as soon as a ClientHello message indicating a re-negotiation is received from the client FortiOS terminates the TCP connection.

Since SSL offloading does not support requesting client certificates the only circumstance in which a renegotiation is required is when more than 2^32 bytes of data are exchanged over a single handshake. If you are sure that this volume of traffic will not occur then you can disable re-negotiation and avoid any possibility of the attack described in CVE-2009-3555.

The re-negotiation behavior can be tested using OpenSSL. The OpenSSL s_client application has the feature that the user can request that it do renegotiation by typing “R”. For example, the following shows a successful renegotiation against a FortiGate unit configured with a VIP for 192.168.2.100:443:

$ openssl s_client -connect 192.168.2.100:443 CONNECTED(00000003)

depth=1 /C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate Authority/CN=support/emailAddress=support@fortinet.com verify error:num=19:self signed certificate in certificate chain verify return:0

—

Certificate chain

0

s:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Fortigate/CN=FW80CM3909604325/emailAdd ress=support@fortinet.com

i:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate

Authority/CN=support/emailAddress=support@fortinet.com

1 s:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate

Authority/CN=support/emailAddress=support@fortinet.com i:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate Authority/CN=support/emailAddress=support@fortinet.com

—

Server certificate

—–BEGIN CERTIFICATE—–

—certificate not shown—

—–END CERTIFICATE—–

subject=/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Fortigate/CN=FW80CM3909604325/em ailAddress=support@fortinet.com

issuer=/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate

Authority/CN=support/emailAddress=support@fortinet.com

—

No client certificate CA names sent

—

SSL handshake has read 2370 bytes and written 316 bytes —

Disabling SSL/TLS re-negotiation

New, TLSv1/SSLv3, Cipher is DHE-RSA-AES256-SHA

Server public key is 1024 bit

Compression: NONE Expansion: NONE SSL-Session:

Protocol : TLSv1

Cipher : DHE-RSA-AES256-SHA Session-ID:

02781E1E368DCCE97A95396FAA82E8F740F5BBA96CF022F6FEC3597B0CC88095

Session-ID-ctx: Master-Key:

A6BBBD8477A2422D56E57C1792A4EA9C86F37D731E67D0A66E5CDB2B5C76650780C0E7F01CFF851EC44661

86F4C48397

Key-Arg : None

Start Time: 1264453027

Timeout : 300 (sec)

Verify return code: 19 (self signed certificate in certificate chain)

—

GET /main.c HTTP/1.0

R

RENEGOTIATING

depth=1 /C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate Authority/CN=support/emailAddress=support@fortinet.com verify error:num=19:self signed certificate in certificate chain verify return:0 HTTP/1.0 200 ok

Content-type: text/plain

/*

* Copyright (C) 2004-2007 Fortinet */

#include <stdio.h> #include “vsd_ui.h”

int main(int argc, char **argv)

{

return vsd_ui_main(argc, argv);

} closed $

The following is the same test, but this time with the VIP configuration changed to ssl-clientrenegotation deny:

$ openssl s_client -connect 192.168.2.100:443 CONNECTED(00000003)

depth=1 /C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate Authority/CN=support/emailAddress=support@fortinet.com verify error:num=19:self signed certificate in certificate chain verify return:0

—

Certificate chain

0

s:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Fortigate/CN=FW80CM3909604325/emailAdd ress=support@fortinet.com

i:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate

Authority/CN=support/emailAddress=support@fortinet.com

1 s:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate

Authority/CN=support/emailAddress=support@fortinet.com i:/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate Authority/CN=support/emailAddress=support@fortinet.com

—

Server certificate

—–BEGIN CERTIFICATE—–

—certificate not shown—

—–END CERTIFICATE—–

subject=/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Fortigate/CN=FW80CM3909604325/em ailAddress=support@fortinet.com

issuer=/C=US/ST=California/L=Sunnyvale/O=Fortinet/OU=Certificate

Authority/CN=support/emailAddress=support@fortinet.com

—

No client certificate CA names sent

—

SSL handshake has read 2370 bytes and written 316 bytes

—

New, TLSv1/SSLv3, Cipher is DHE-RSA-AES256-SHA

Server public key is 1024 bit

Compression: NONE Expansion: NONE SSL-Session:

Protocol : TLSv1

Cipher : DHE-RSA-AES256-SHA Session-ID:

8253331D266DDE38E4D8A04AFCA9CBDED5B1134932CE1718EED6469C1FBC7474

Session-ID-ctx: Master-Key:

ED05A3EF168AF2D06A486362FE91F1D6CAA55CEFC38A3C36FB8BD74236BF2657D4701B6C1456CEB5BB5EFA

A7619EF12D

Key-Arg : None

Start Time: 1264452957

Timeout : 300 (sec)

Verify return code: 19 (self signed certificate in certificate chain)

—

GET /main.c HTTP/1.0

R

RENEGOTIATING

19916:error:1409E0E5:SSL routines:SSL3_WRITE_BYTES:ssl handshake failure:s3_pkt.c:530: Use the following command to check the SSL stats to see that the renegotiations blocked counter is now 1:

diagnose firewall vip virtual-server stats ssl ssl client connections total 0 active 0 max 0

handshakes total 4 active 0 max 0 completed 4 abbreviated 0 session states total 4 active 4 max 4

cipher-suite failures 0

embryonics total 0 active 0 max 0 terminated 0

Disabling SSL/TLS re-negotiation

renegotiations blocked 1

server connections total 0 active 0 max 0

handshakes total 3 active 0 max 0 completed 2 abbreviated 1 session states total 1 active 1 max 1

cipher-suite failures 0

internal error 0 bad handshake length 0 bad change cipher spec length 0 pubkey too big 0 persistence

find 0 found 0 clash 0 addr 0 error 0

If the virtual server debug log is examined (diagnose debug appl vs -1) then at the point the re-negotiation is blocked there is a log:

vs ssl 12 handshake recv ClientHello vs ssl 12 handshake recv 1

(0100005403014b5e056c7f573a563bebe0258c3254bbaff7046a461164f34f94f4f3d019c418000026

00390038003500160013000a00330032002f00050004001500120009001400110008000600030201000

00400230000) vs ssl 12 client renegotiation attempted rejected, abort

vs ssl 12 closing 0 up vs src 12 close 0 in vs src 12 error closing vs dst 14 error closing vs dst 14 closed vs ssl 14 close vs sock 14 free vs src 12 closed vs ssl 12 close vs sock 12 free

HTTP and HTTPS load balancing, multiplexing, and persistence

In a firewall load balancing virtual server configuration, you can select HTTP to load balance only HTTP sessions. The virtual server will load balance HTTP sessions received at the virtual server interface with destination IP address that matches the configured virtual server IP and destination port number that matches the configured virtual server port. The default virtual server port for HTTP load balancing is 80, but you can change this to any port number. Similarly for HTTPS load balancing, set the virtual server type to HTTPS and then select the interface, virtual server IP, and virtual server port that matches the HTTPS traffic to be load balanced. Usually HTTPS traffic uses port 443.

You can also configure load balancing to offload SSL processing for HTTPS and SSL traffic. See SSL/TLS load balancing on page 36.

HTTP and HTTPS multiplexing

For both HTTP and HTTPS load balancing you can multiplex HTTP requests and responses over a single TCP connection. HTTP multiplexing is a performance saving feature of HTTP/1.1 compliant web servers that provides the ability to pipeline many unrelated HTTP or HTTPS requests on the same connection. This allows a single HTTPD process on the server to interleave and serve multiple requests. The result is fewer idle sessions on the web server so server resources are used more efficiently. HTTP multiplexing can take multiple separate inbound sessions and multiplex them over the same internal session. This may reduce the load on the backend server and increase the overall performance.

HTTP multiplexing may improve performance in some cases. For example, if users web browsers are only compatible with HTTP 1.0. HTTP multiplexing can also improve performance between a web server and the FortiGate unit if the FortiGate unit is performing SSL acceleration. However, in most cases HTTP multiplexing should only be used if enabling it leads to a measurable improvement in performance.

To enable HTTP multiplexing from the GUI, select multiplex HTTP requests/responses over a single TCP connection. To enable HTTP multiplexing from the CLI enable the http-multiplex option.

Preserving the client IP address

Select preserve client IP from the GUI or enable the http-ip-header option from the CLI to preserve the IP address of the client in the X-Forwarded-For HTTP header. This can be useful in an HTTP multiplexing configuration if you want to be able to see the original client IP address in log messages on the destination web server. If this option is not selected, the X-Forwarded-For HTTP header contains the IP address of the FortiGate unit.

Preserving the client IP address in a different HTTP header

If you select preserve client IP from the GUI or enable the http-ip-header option from the CLI you can also preserve the client IP in a different HTTP header. This can be useful if you want to use a custom header name instead of X-Forwarded-For.

You can add the custom header name from the CLI. When http-ip-header is enabled you can add a custom header name to the http-ip-header-name option. If you don’t add a name the X-Forwarded-For header is used.

HTTP and HTTPS persistence

Configure load balancing persistence for HTTP or HTTPS to make sure that a user is connected to the same server every time they make a request that is part of the same session. HTTP cookie persistence uses injected cookies to enable persistence.

When you configure persistence, the FortiGate unit load balances a new session to a real server according to the Load Balance Method. If the session has an HTTP cookie or an SSL session ID, the FortiGate unit sends all subsequent sessions with the same HTTP cookie or SSL session ID to the same real server.

The following example shows how to enable cookie persistence and set the cookie domain to .example.org.

config firewall vip edit HTTP_Load_Balance set type server-load-balance set server-type http set extport 8080 set extintf port2 set extip 192.168.20.20 set persistence http-cookie set http-cookie-domain .example.org config realservers edit 1 set ip 10.10.10.1

set port 80

next edit 2 set ip 10.10.10.2

set port 80

next edit 3 set ip 10.10.10.3

set port 80

end

How HTTP cookie persistence options work

The following options are available for the config firewall vip command when type is set to serverload-balance, server-type is set to http or https and persistence is set to http-cookie:

http-cookie-domain-from-host http-cookie-domain http-cookie-path http-cookie-generation http-cookie-age http-cookie-share https-cookie-share

When HTTP cookie persistence is enabled the FortiGate unit inserts a header of the following form into each HTTP response unless the corresponding HTTP request already contains a FGTServer cookie:

Set-Cookie: FGTServer=E7D01637C4B08E89A6714213A9D85D9C7E4D8158; Version=1; Max-Age=3600

HTTP and HTTPS

The value of the FGTServer cookie encodes the server that traffic should be directed to. The value is encoded so as to not leak information about the internal network.

Enable http-cookie-domain-from-host to extract the cookie domain from the host: header in the HTTP request. For example, to restrict the cookie to.server.com, enter:

The generated cookies could have the following form if the Host: header contains exhost.com:

Set-Cookie: FGTServer=E7D01637C4B08E89A6714213A9D85D9C7E4D8158; Version=1; Domain=.exhost.com; Max-Age=3600

For more information, see “HTTP host-based load balancing”.

Use http-cookie-domain to restrict the domain that the cookie should apply to. For example, to restrict the cookie to.server.com, enter:

set http-cookie-domain .server.com

Now all generated cookies will have the following form:

Set-Cookie: FGTServer=E7D01637C4B08E89A6714213A9D85D9C7E4D8158; Version=1; Domain=.server.com; Max-Age=3600

Use http-cookie-path to limit the cookies to a particular path. For example, to limit cookies to the path /sales, enter:

set http-cookie-path /sales

Now all generated cookies will have the following form:

Set-Cookie: FGTServer=E7D01637C4B08E89A6714213A9D85D9C7E4D8158; Version=1; Domain=.server.com; Path=/sales; Max-Age=3600

Use http-cookie-age to change how long the browser caches the cookie. You can enter an age in minutes or set the age to 0 to make the browser keep the cookie indefinitely:

set http-cookie-age 0

Now all generated cookies will have the following form:

Set-Cookie: FGTServer=E7D01637C4B08E89A6714213A9D85D9C7E4D8158; Version=1; Domain=.server.com; Path=/sales

Use http-cookie-generation to invalidate all cookies that have already been generated. The exact value of the generation is not important, only that it is different from any generation that has already been used for cookies in this domain. The simplest approach is to increment the generation by one each time invalidation is required. Since the default is 0, enter the following to invalidate all existing cookies:

set http-cookie-generation 1

Use http-cookie-share {disable | same-ip} to control the sharing of cookies across virtual servers in the same virtual domain. The default setting same-ip means that any FGTServer cookie generated by one virtual server can be used by another virtual server in the same virtual domain. For example, if you have an application that starts on HTTP and then changes to HTTPS and you want to make sure that the same server is used for the HTTP and HTTPS traffic then you can create two virtual servers, one for port 80 (for HTTP) and one for port 443 (for HTTPS). As long as you add the same real servers to both of these virtual servers (and as long as both virtual servers have the same number of real servers with the same IP addresses), then cookies generated by accessing the HTTP server are reused when the application changes to the HTTPS server.

If for any reason you do not want this sharing to occur then select disable to make sure that a cookie generated for a virtual server cannot be used by other virtual servers.

Use https-cookie-secure to enable or disable using secure cookies. Secure cookies are disabled by default because secure cookies can interfere with cookie sharing across HTTP and HTTPS virtual servers. If enabled, then the Secure tag is added to the cookie inserted by the FortiGate unit:

Set-Cookie: FGTServer=E7D01637C4B08E89A6714213A9D85D9C7E4D8158; Version=1; Max-Age=3600; Secure

HTTP host-based load balancing

When configuring HTTP or HTTPS load balancing you can select HTTP host load balancing to load balance HTTP host connections across multiple real servers using the host’s HTTP header to guide the connection to the correct real server. HTTP 1.1 includes the concept of a virtual server which allows a HTTP or HTTPS server with a single external IP address to serve requests for multiple DNS domains by using the mandatory Host: header in a HTTP request to indicate which DNS domain the request is destined for.

FortiOS can load-balance HTTP and HTTPS connections among multiple real servers using the Host: header to guide the connection to the correct real server. The host load balancing method allows a real server to specify a http-host attribute which is the domain name of the traffic for that real server. Each real server can only specify a single domain name. The same domain name can appear in more than one real server but only the first one that is up will be used, any others are purely for redundancy. If the Host: header contains a domain that does not match any http-host entry then the connection will be dropped. A real server with no http-host can be matched by any Host: domain.

For example, consider a FortiGate unit that is load-balancing traffic to three real servers. Traffic for www.example1.com should go to 192.168.2.1, traffic for www.example2.com should go to 192.168.2.2 and traffic to any other domain should go to 192.168.2.3. To enable this configuration you would add a virtual server and set the load balance method to HTTP host. Then you would add three real servers and set the HTTP host of the real server with IP address 192.168.2.1 to www.example1.com, the HTTP host of the real server with IP address 192.168.2.2 to www.example2.com and you would not specify an HTTP host for the third real server.

The configuration of a virtual IP to achieve this result would be:

config firewall vip edit “http-host-ldb” set type server-load-balance set extip 172.16.67.195 set extintf “lan” set server-type http set ldb-method http-host set extport 80 config realservers

edit 1 set http-host “www.example1.com” set ip 192.168.2.1

set port 80

next edit 2 set http-host “www.example2.com” set ip 192.168.2.2

set port 80

next edit 3 set ip 192.168.2.3

set port 80

next end end

Host load balancing and HTTP cookie persistence

In an HTTP host-based load balancing configuration with HTTP cookie persistence enabled you can optionally configure cookie persistence to use the domain set in the host header as the cookie domain. You can do this by enabling the http-cookie-domain-from-host option, for example:

config firewall vip edit “http-host-ldb” set type server-load-balance set extip 172.16.67.195 set extintf “lan” set server-type http set ldb-method http-host set extport 80 set persistence http-cookie set http-cookie-domain-from-host enable config realservers edit 1 set http-host “www.example1.com” set ip 192.168.2.1

set port 80 next edit 2 set http-host “www.example2.com” set ip 192.168.2.2

set port 80 next edit 3 set ip 192.168.2.3

set port 80 next end

end

Configuring load balancing

This section describes how to use the FortiOS server load balancing to load balance traffic to multiple backend servers.

You can configure FortiOS load balancing to intercept incoming traffic with a virtual server and distribute it among one or more backend real servers. By doing so, FortiOS enables multiple real servers to respond as if they were a single device or virtual server. This in turn means that more simultaneous requests can be handled by the servers.

Traffic can be balanced across multiple backend real servers based on a selection of load balancing methods including static (failover), round robin, weighted to account for different sized servers, or based on the health and performance of the server including round trip time, number of connections. The load balancer can balance layer 7 HTTP, HTTPS, SSL, generic layer 4 TCP, UDP and generic layer 3 IP protocols. Session persistence is supported based on injected HTTP/HTTPS cookies or the SSL session ID.

You can bind up to 8 real servers can to one virtual server. The real server topology is transparent to end users, and the users interact with the system as if it were only a single server with the IP address and port number of the virtual server. The real servers may be interconnected by high-speed LAN or by geographically dispersed WAN. The FortiGate unit schedules requests to the real servers and makes parallel services of the virtual server to appear to involve a single IP address.

There are additional benefits to load balancing. First, because the load is distributed across multiple servers, the service being provided can be highly available. If one of the servers breaks down, the load can still be handled by the other servers. Secondly, this increases scalability. If the load increases substantially, more servers can be added behind the FortiGate unit to cope with the increased load.

Server load balancing configuration

Traffic can be balanced across multiple backend real servers based on a selection of load balancing methods including static (failover), round robin, weighted to account for different sized servers, or based on the health and performance of the server including round trip time, number of connections. The load balancer can balance layer 7 HTTP, HTTPS, SSL, generic layer 4 TCP, UDP and generic layer 3 IP protocols. Session persistence is supported based on injected HTTP/HTTPS cookies or the SSL session ID.

You can bind up to 8 real servers can to one virtual server. The real server topology is transparent to end users, and the users interact with the system as if it were only a single server with the IP address and port number of the Load balancing and other FortiOS features

virtual server. The real servers may be interconnected by high-speed LAN or by geographically dispersed WAN. The FortiGate unit schedules requests to the real servers and makes parallel services of the virtual server to appear to involve a single IP address.

Load balancing and other FortiOS features

Flow-based and proxy-based security features such as virus scanning, IPS, DLP, application control, and web filtering can be applied to load balanced sessions. This includes SSL offloading and multiplexing. Applying these UTM features to load balancing traffic may reduce load balancing performance.

Authentication is not supported for load balancing sessions. Usually FortiGate load balancing is used to allow public access to services on servers protected by a FortiGate unit. Authentication is not generally not required for this kind of configuration.

Features such web proxying, web caching, and WAN optimization also do not work with load balanced sessions. However, most other features that can be applied by a security policy are supported.

Configuring load balancing from the GUI

A virtual server is a specialized firewall virtual IP that performs server load balancing. From the GUI you add load balancing virtual server by going to Policy & Objects > Virtual Servers.

You can use the GUI to configure IPv, IPv6, IPv4 to IPv6 (NAT46), or IPv6 to IPv4 (NAT64) load balancing.

Type

Select the type of virtual server to configure. You can select IPv4, IPv6, NAT46, or NAT64. If Type is set to NAT46 or NAT64 you have fewer load balancing options (just HTTP, TCP, UDP and IP) and you can’t configure advanced SSL and HTTPS load balancing features.

Name

Enter the name for the virtual server.

Type

Select the protocol to be load balanced by the virtual server. If you select a general protocol such as IP, TCP, or

UDP the virtual server load balances all IP, TCP, or UDP sessions. If you select specific protocols such as HTTP, HTTPS, or SSL you can apply additional server load balancing features such as Persistence and HTTP Multiplexing.

• Select HTTP to load balance only HTTP sessions with destination port number that matches the Virtual Server

Port setting. Change Virtual Server Port to match the destination port of the sessions to be load balanced (usually port 80 for HTTP sessions). You can also select HTTP Multiplex. You can also set Persistence to HTTP Cookie to select cookie-based persistence.

• Select HTTPS to load balance only HTTPS sessions with destination port number that matches the Virtual Server

Port setting. Change Virtual Server Port to match the destination port of the sessions to be load balanced

(usually port 443 for HTTPS sessions). You can also select Multiplex HTTP requests/responses. You can also set Persistence to HTTP Cookie to select cookie-based persistence. You can also set Persistence to SSL Session ID.

• Select IMAPS to load balance only IMAPS sessions with destination port number that matches the Virtual Server Port Change Virtual Server Port to match the destination port of the sessions to be load balanced (usually port 993 for IMAPS sessions). You can also set Persistence to SSL Session ID. l Select POP3S to load balance only POP3S sessions with destination port number that matches the Virtual Server Port setting. Change Virtual Server Port to match the destination port of the sessions to be load balanced (usually port 995 for POP3S sessions). You can also set Persistence to SSL Session ID.
• Select SMTPS to load balance only SMTPS sessions with destination port number that matches the Virtual Server Port Change Virtual Server Port to match the destination port of the sessions to be load balanced (usually port 465 for SMTPS sessions). You can also set Persistence to SSL Session ID.
• Select SSL to load balance only SSL sessions with destination port number that matches the Virtual Server Port Change Virtual Server Port to match the destination port of the sessions to be load balanced.
• Select TCP to load balance only TCP sessions with destination port number that matches the Virtual Server Port Change Virtual Server Port to match the destination port of the sessions to be load balanced.
• Select UDP to load balance only UDP sessions with destination port number that matches the Virtual Server Port Change Virtual Server Port to match the destination port of the sessions to be load balanced. l Select IP to load balance all sessions accepted by the security policy that contains this virtual server.

Interface

Select the virtual server external or outgoing interface from the list. The outgoing interface is connected to the source network and receives the packets to be forwarded to the destination network.

Virtual Server IP

The IPv4 address of the virtual server. This is an IP address on the external interface that you want to map to an address on the destination network.

Virtual Server Port

Enter the external port number that you want to map to a port number on the destination network. Sessions with this destination port are load balanced by this virtual server.

Load Balance Method

Select the load balancing method used by the virtual server.

Persistence

Configure persistence to make sure that a user is connected to the same server every time they make a request that is part of the same session. Session persistence is supported for HTTP and SSL sessions.

Health Check

Select which health check monitor configuration will be used to determine a server’s connectivity status.

HTTP Multiplexing

Select to use the FortiGate unit to multiplex multiple client connections into a few connections between the FortiGate unit and the real server.

Preserve Client IP

Select to preserve the IP address of the client in the X-Forwarded-For HTTP header. This can be useful if you want log messages on the real servers to the client’s original IP address. If this option is not selected, the header will contain the IP address of the FortiGate unit.

This option appears only if Type is set to HTTP or HTTPS.

SSL Offloading

Accelerate clients’ SSL connections to the server by using the FortiGate to perform SSL operations. This option appears only if Type is set to one of the SSL protocols.

Mode

Select which segments of the SSL connection will receive SSL offloading. You can select Client <-> FortiGate (or half mode) or Full (full mode).

This option appears only if Type is set to one of the SSL protocols.

Certificate

Select the certificate to use with SSL Offloading. The certificate key size must be 1024 or 2048 bits. 4096-bit keys are not supported.

This option appears only if Type is set to one of the SSL protocols.

Real Servers

Add Real Servers to the virtual server. The virtual server load balances traffic to these real servers. See Real servers on page 23.

Configuring load balancing from the CLI

From the CLI you configure IPv4 load balancing by adding a firewall virtual IP and setting the virtual IP type to server load balance:

config firewall vip edit Vserver-HTTP-1 set type server-load-balance …

Sever load balancing is also supported for:

Load balancing methods

l IPv6 using the command config firewall vip6 l IPv6 to IPv4 using the command config firewall vip64 l IPv4 to IPv6 using the commmand config firewall vip46

Configuration is the same as IPv4 VIPs except support for advanced HTTP and SSL related features is not available. IPv6 server load balancing supports all the same server types as IPv4 server load balancing (HTTP, HTTPS, IMAPS, POP3S, SMTPS, SSL, TCP, UDP, and IP). IPv4 to IPv6 and IPv6 to IPv4 server load balancing supports fewer server types (HTTP, TCP, UDP, and IP).

A virtual server includes a virtual server IP address bound to an interface. The virtual server IP address is the destination address incoming packets to be load balanced and the virtual server is bound to the interface that receives the packets to be load balanced.

For example, if you want to load balance incoming HTTP traffic from the Internet to a group of web servers on a DMZ network, the virtual server IP address is the known Internet IP address of the web servers and the virtual server binds this IP address to the FortiGate interface connected to the Internet.

When you bind the virtual server’s external IP address to a FortiGate unit interface, by default, the network interface responds to ARP requests for the bound IP address. Virtual servers use proxy ARP, as defined in RFC 1027, so that the FortiGate unit can respond to ARP requests on a network for a real server that is actually installed on another network. In some cases you may not want the network interface sending ARP replies. You can use the arp-reply option disable sending ARP replies:

config firewall vip edit Vserver-HTTP-1 set type server-load-balance set arp-reply disable …

The load balancing virtual server configuration also includes the virtual server port. This is the TCP port on the bound interface that the virtual server listens for traffic to be load balanced on. The virtual server can listen on any port.

Load balancing methods

The load balancing method defines how sessions are load balanced to real servers. A number of load balancing methods are available as listed below.

All load balancing methods will not send traffic to real servers that are down or not responding. However, the FortiGate unit can only determine if a real server is not responding by using a health check monitor. You should always add at least one health check monitor to a virtual server or to individual real servers, or load balancing methods may attempt to distribute sessions to real servers that are not functioning.

Static

The traffic load is statically spread evenly across all real servers. However, sessions are not assigned according to how busy individual real servers are. This load balancing method provides some persistence because all sessions from the same source address always go to the same real server. However, the distribution is stateless, so if a real server is added or removed (or goes up or down) the distribution is changed and persistence could be lost.

Session persistence

Round Robin

Directs new requests to the next real server, and treats all real servers as equals regardless of response time or number of connections. Dead real servers or non responsive real servers are avoided.

Weighted

Real servers with a higher weight value receive a larger percentage of connections. Set the real server weight when adding a real server.

Least Session

Directs requests to the real server that has the least number of current connections. This method works best in environments where the real servers or other equipment you are load balancing all have similar capabilities. This load balancing method uses the FortiGate session table to track the number of sessions being processed by each real server. The FortiGate unit cannot detect the number of sessions actually being processed by a real server.

Least RTT

Directs sessions to the real server with the least round trip time. The round trip time is determined by a Ping health check monitor and is defaulted to 0 if no Ping health check monitors are added to the virtual server.

First Alive

Always directs sessions to the first alive real server. This load balancing schedule provides real server failover protection by sending all sessions to the first alive real server and if that real server fails, sending all sessions to the next alive real server. Sessions are not distributed to all real servers so all sessions are processed by the “first” real server only.

First refers to the order of the real servers in the virtual server configuration. For example, if you add real servers

A, B and C in that order, then all sessions always go to A as long as it is alive. If A goes down then sessions go to B and if B goes down sessions go to C. If A comes back up sessions go back to A. Real servers are ordered in the virtual server configuration in the order in which you add them, with the most recently added real server last. If you want to change the order you must delete and re-add real servers in the required order.

HTTP Host

Load balances HTTP host connections across multiple real servers using the host’s HTTP header to guide the connection to the correct real server.

Session persistence

Use persistence to make sure that a user is connected to the same real server every time they make an HTTP,

HTTPS, or SSL request that is part of the same user session. For example, if you are load balancing HTTP and HTTPS sessions to a collection of eCommerce web servers, when a user is making a purchase they will be starting multiple sessions as they navigate the eCommerce site. In most cases all of the sessions started by this user during on eCommerce session should be processed by the same real server. Typically, the HTTP protocol Real servers

keeps track of these related sessions using cookies. HTTP cookie persistence makes sure that all sessions that are part of the same user session are processed by the same real server

When you configure persistence, the FortiGate unit load balances a new session to a real server according to the load balance method. If the session has an HTTP cookie or an SSL session ID, the FortiGate unit sends all subsequent sessions with the same HTTP cookie or SSL session ID to the same real server. For more information about HTTP and HTTPS persistence, see “HTTP and HTTPS persistence”.

Real servers

Add real servers to a load balancing virtual server to provide the information the virtual server requires to be able to send sessions to the server. A real server configuration includes the IP address of the real server and port number that the real server receives sessions on. The FortiGate unit sends sessions to the real server’s IP address using the destination port number in the real server configuration.

When configuring a real server you can also specify the weight (used if the load balance method is set to weighted) and you can limit the maximum number of open connections between the FortiGate unit and the real server. If the maximum number of connections is reached for the real server, the FortiGate unit will automatically switch all further connection requests other real servers until the connection number drops below the specified limit. Setting Maximum Connections to 0 means that the FortiGate unit does not limit the number of connections to the real server.

Real server active, standby, and disabled modes

By default the real server mode setting is active indicating that the real server is available to receive connections. If the real server is removed from the network (for example, for routine maintenance or because of a hardware or software failure) you can change the mode to standby or disabled. In disabled mode the FortiGate unit no longer sends sessions to the real server.

If a real server is in standby mode the FortiGate also does not send sessions to it unless other real servers added to the same virtual server become unavailable. For example:

• A virtual server that includes two real servers one in active mode and one in standby mode. If the real server in active mode fails, the real server in standby mode is changed to active mode and all sessions are sent to this real server.
• A virtual server includes three real servers, two in active mode and one in standby mode, if one of the real servers in active mode fails, the real server in standby mode is changed to active mode and sessions are load balanced between it and still operating real server. If both real servers in active mode fail, all sessions are sent to the real server in standby mode.

Adding real servers from the GUI

To add a real server from the GUI go to Policy & Objects > Virtual Servers, edit a virtual server and under Real Servers select Create New to add a real server to this virtual server.

IP Address

Enter the IP address of the real server.

Real servers

Port

Enter the port number on the destination network to which the external port number is mapped.

Weight

Enter the weight value of the real server. The higher the weight value, the higher the percentage of connections the server will handle. A range of 1-255 can be used. This option is available only if the associated virtual server’s load balance method is Weighted.

Max Connections

Enter the limit on the number of active connections directed to a real server. A range of 1-99999 can be used. If the maximum number of connections is reached for the real server, the FortiGate unit will automatically switch all further connection requests to another server until the connection number drops below the specified limit.

Setting Maximum Connections to 0 means that the FortiGate unit does not limit the number of connections to the real server.

HTTP Host

Enter the HTTP header for load balancing across multiple real servers. This feature is used for load balancing HTTP host connections across multiple real servers using the host’s HTTP header to guide the connection to the correct real server, providing better load balancing for those specific connections.

Mode

Select a mode for the real server. The real server can be active, on standby, or disabled.

Adding real servers from the CLI

To add a real server from the CLI you configure a virtual server and add real servers to it. For example, to add three real servers to a virtual server that load balances UDP sessions on port 8190 using weighted load balancing. For each real server the port is not changed. The default real server port is 0 resulting in the traffic being sent the real server with destination port 8190. Each real sever is given a different weight. Servers with higher weights have a max-connections limit to prevent too many sessions from being sent to them.

config firewall vip edit Vserver-UDP-1 set type server-load-balance set server-type udp set ldb-method weighted set extip 172.20.120.30 set extintf wan1 set extport 8190 set monitor ping-mon-1 config realservers edit 1 set ip 10.31.101.30 set weight 100 set max-connections 10000

next edit 2 set ip 10.31.101.40 set weight 100

Health check monitoring

set max-connections 10000

next edit 3 set ip 10.31.101.50 set weight 10

end

end

Health check monitoring

From the FortiGate GUI you can go to Policy & Objects > Health Check and configure health check monitoring so that the FortiGate unit can verify that real servers are able respond to network connection attempts. If a real server responds to connection attempts the load balancer continues to send sessions to it. If a real server stops responding to connection attempts the load balancer assumes that the server is down and does not send sessions to it. The health check monitor configuration determines how the load balancer tests the real servers. You can use a single health check monitor for multiple load balancing configurations.

You can configure TCP, HTTP and Ping health check monitors. Usually you would want the health check monitor to use the same protocol for checking the health of the server as the traffic being load balanced to it. For example, for an HTTP load balancing configuration you would normally use an HTTP health check monitor.

For the TCP and HTTP health check monitors you can specify the destination port to use to connect to the real servers. If you set the port to 0, the health check monitor uses the port defined in the real server. This allows you to use the same health check monitor for multiple real servers using different ports. You can also configure the interval, timeout and retry. A health check occurs every number of seconds indicated by the interval. If a reply is not received within the timeout period the health check is repeated every second. If no response is received after the number of configured retires, the virtual server is considered unresponsive, and load balancing does not srend traffic to that real server. The health check monitor will continue to contact the real server and if successful, the load balancer can resume sending sessions to the recovered real server.

The default health check configuration has an interval of 10 seconds, a timeout of 2 seconds and a retry of 3. This means that the health check monitor checks the health of a real server every 10 seconds. If a reply is not received within 2 seconds the health check monitor re-checks the server every second for 3 retries. If no response is received for 2 seconds after the final retry the server is considered unresponsive. This entire process takes a total of 7 seconds to consider a virtual server as unresponsive (2 second timeout + (3 re-checks x 1 second) + 2 second timeout = 7 seconds). Since this health check process is repeated every 10 seconds, a server can be down for a maximum of 10 + 7 = 17 seconds before the health check monitor considers it down.

For HTTP health check monitors, you can add URL that the FortiGate unit connects to when sending a get request to check the health of a HTTP server. The URL should match an actual URL for the real HTTP servers. The URL is optional.

The URL would not usually include an IP address or domain name. Instead it should start with a “/” and be followed by the address of an actual web page on the real server. For example, if the IP address of the real server is 10.31.101.30, the URL “/test_page.htm” causes the FortiGate unit to send an HTTP get request to “http://10.31.101.30/test_page.htm”.

For HTTP health check monitors, you can also add a matched content phrase that a real HTTP server should include in response to the get request sent by the FortiGate unit using the content of the URL option. If the URL returns a web page, the matched content should exactly match some of the text on the web page. You can use the URL and Matched Content options to verify that an HTTP server is actually operating correctly by responding to get requests with expected web pages. Matched content is only required if you add a URL.

Health check monitoring

For example, you can set matched content to “server test page” if the real HTTP server page defined by the URL option contains the phrase “server test page”. When the FortiGate unit receives the web page in response to the URL get request, the system searches the content of the web page for the matched content phrase.

Name

Enter the name of the health check monitor configuration.

Type

Select the protocol used to perform the health check.

l TCP l HTTP l PING

Port

Enter the port number used to perform the health check. If you set the Port to 0, the health check monitor uses the port defined in the real server. This way you can use a single health check monitor for different real servers.

This option does not appear if the Type is PING.

Interval

Enter the number of seconds between each server health check.

URL

For HTTP health check monitors, add a URL that the FortiGate unit uses when sending a get request to check the health of a HTTP server. The URL should match an actual URL for the real HTTP servers. The URL is optional.

The URL would not usually include an IP address or domain name. Instead it should start with a “/” and be followed by the address of an actual web page on the real server. For example, if the IP address of the real server is 10.10.10.1, the URL “/test_page.htm” causes the FortiGate unit to send an HTTP get request to “http://10.10.10.1/test_page.htm”.

This option appears only if Type is HTTP.

Matched Content

For HTTP health check monitors, add a phrase that a real HTTP server should include in response to the get request sent by the FortiGate unit using the content of the URL option. If the URL returns a web page, the Matched Content should exactly match some of the text on the web page. You can use the URL and Matched Content options to verify that an HTTP server is actually operating correctly by responding to get requests with expected web pages. Matched content is only required if you add a URL.

For example, you can set Matched Content to “server test page” if the real HTTP server page defined by the URL option contains the phrase “server test page”. When the FortiGate unit receives the web page in response to the URL get request, the system searches the content of the web page for the Matched Content phrase.

This option appears only if Type is HTTP.

Load balancing limitations

Max Redirects

For an HTTP health check monitor, specify the maximum number of redirects that the health check monitor will follow when testing the health of the real HTTP server. This feature allows you to do health checking of the HTTP server is accessed through one or more redirects.

Timeout

Enter the number of seconds which must pass after the server health check to indicate a failed health check.

Retry

Enter the number of times, if any, a failed health check will be retried before the server is determined to be inaccessible.

Load balancing limitations

The following limitations apply when adding virtual IPs, load balancing virtual servers, and load balancing real servers. Load balancing virtual servers are actually server load balancing virtual IPs. You can add server load balance virtual IPs from the CLI.

• Virtual IP External IP Address/Range entries or ranges cannot overlap with each other or with load balancing virtual server Virtual Server IP
• A virtual IP Mapped IP Address/Range cannot be 0.0.0.0 or 255.255.255.255. l A real server IP cannot be 0.0.0.0 or 255.255.255.255.
• If a static NAT virtual IP External IP Address/Range is 0.0.0.0, the Mapped IP Address/Range must be a single IP address. l If a load balance virtual IP External IP Address/Range is 0.0.0.0, the Mapped IP Address/Range can be an address range.
• When port forwarding, the count of mapped port numbers and external port numbers must be the same. The GUI does this automatically but the CLI does not. l Virtual IP and virtual server names must be different from firewall address or address group names.

Monitoring load balancing

From the GUI you can go to Monitor > Load Balance Monitor to monitor the status of configured virtual servers and real servers and start or stop the real servers. You can also use the get test ipldb command from the CLI to display similar information.

For each real server the monitor displays health status (up or down), active sessions, round trip time (RTT) and the amount of bytes of data processed. From the monitor page you can also stop sending new sessions to any real server. When you select to stop sending sessions the FortiGate unit performs of graceful stop by continuing to send data for sessions that were established or persistent before you selected stop. However, no new sessions are started.

Real Server

The IP addresses of the existing real servers.

Status

Displays the health status according to the health check results for each real server. A green arrow means the server is up. A red arrow means the server is down.

Mode

The mode of the health check monitor. Can be active, standby, or disabled.

Monitor Events

Display each real server’s up and down times.

Active Sessions

Display each real server’s active sessions.

RTT (ms)

Displays the Round Trip Time (RTT) of each real server. By default, the RTT is “<1”. This value will change only when ping monitoring is enabled on a real server.

Bytes Processed

Displays the traffic processed by each real server.

Graceful Stop/Start

Select to start or stop real servers. When stopping a server, the FortiGate unit will not accept new sessions but will wait for the active sessions to finish.

Load balancing diagnose commands

You can also use the following diagnose commands to view status information for load balancing virtual servers and real servers:

diagnose firewall vip realserver {down | healthcheck | list | up} diagnose firewall vip virtual-server {filter | real-server | stats}

For example, the following command lists and displays status information for all real servers: diagnose firewall vip virtual-server real-server

vd root/0 vs vs/2 addr 10.31.101.30:80 status 1/1 conn: max 0 active 0 attempts 0 success 0 drop 0 fail 0

Load balancing diagnose commands

vd root/0 vs vs/2 addr 10.31.101.20:80 status 1/1 conn: max 0 active 0 attempts 0 success 0 drop 0 fail 0

Many of the diagnostic commands involve retrieving information about one or more virtual servers. To control which servers are queried you can define a filter:

diagnose firewall vip virtual-server filter <filter_str> Where <filter_str> can be:

l clear erase the current filter l dst the destination address range to filter by l dst-port the destination port range to filter by l list display the current filter l name the vip name to filter by l negate negate the specified filter parameter l src the source address range to filter by l src-port the source port range to filter by l vd index of virtual domain. -1 matches all

The default filter is empty so no filtering is done.

Logging diagnostics

The logging diagnostics provide information about two separate features:

diagnose firewall vip virtual-server filter

l filter sets a filter for the virtual server debug log l The filter option controls what entries the virtual server daemon will log to the console if diagnose debug application vs level is non-zero. The filtering can be done on source, destination, virtual-server name, virtual domain, and so on:

diagnose firewall vip virtual-server filter <filter_str> Where <filter_str> can be l clear erase the current filter l dst the destination address range to filter by l dst-port the destination port range to filter by l list display the current filter l name the virtual-server name to filter by l negate negate the specified filter parameter l src the source address range to filter by l src-port the source port range to filter by l vd index of virtual domain. -1 matches all

The default filter is empty so no filtering is done.

Real server diagnostics

Enter the following command to list all the real servers:

diagnose firewall vip virtual-server real-server list

In the following example there is only one virtual server called slb and it has two real-servers:

diagnose firewall vip virtual-server server

Load balancing diagnose commands

vd root/0 vs slb/2 addr 172.16.67.191:80 status 1/1 conn: max 10 active 0 attempts 0 success 0 drop 0 fail 0 http: available 0 total 0

vd root/0 vs slb/2 addr 172.16.67.192:80 status 1/1 conn: max 10 active 1 attempts 4 success 4 drop 0 fail 0 http: available 1 total 1

The status indicates the administrative and operational status of the real-server.

• max indicates that the real-server will only allow 10 concurrent connections.
• active is the number of current connections to the server attempts is the total number of connections attempted success is the total number of connections that were successful.
• drop is the total number of connections that were dropped because the active count hit max.
• fail is the total number of connections that failed to complete due to some internal problem (for example, lack of memory).

If the virtual server has HTTP multiplexing enabled then the HTTP section indicates how many established connections to the real-sever are available to service a HTTP request and also the total number of connections.

Basic load balancing configuration example

This section describes the steps required to configure the load balancing configuration shown below. In this configuration a FortiGate-51B unit is load balancing HTTP traffic from the Internet to three HTTP servers on the Internal network. HTTP sessions are accepted at the wan1 interface with destination IP address 172.20.120.121 on TCP port 8080 and forwarded from the internal interface to the web servers. When forwarded the destination address of the sessions is translated to the IP address of one of the web servers.

The load balancing configuration also includes session persistence using HTTP cookies, round-robin load balancing, and TCP health monitoring for the real servers. Ping health monitoring consists of the FortiGate unit using ICMP ping to make sure the web servers can respond to network traffic.

Virtual server and real servers setup

To configure the example load balancing configuration – general configuration steps

1. Add a load balance ping health check monitor.

A ping health check monitor causes the FortiGate unit to ping the real servers every 10 seconds. If one of the servers does not respond within 2 seconds, the FortiGate unit will retry the ping 3 times before assuming that the HTTP server is not responding.

2. Add a load balance virtual server.
3. Add the three load balance real servers to the virtual server.
4. Add a security policy that includes the load balance virtual server as the destination address.

To configure the example load balancing configuration

1. Go to Policy & Objects > Health Check and add the following health check monitor.

2. Go to Policy & Objects > Virtual Servers and add a virtual server that accepts the traffic to be load balanced.

3. On the same GUI page and the real servers to the virtual server.

Basic load balancing configuration example

4. Go to Policy & Objects > IPv4 Policy and add a wan1 to internal security policy that includes the virtual server.

This policy also applies an Antivirus profile to the load balanced sessions.

5. Select OK.

To configure the example load balancing configuration from the CLI

1. Use the following command to add a Ping health check monitor.

config firewall ldb-monitor edit ping-mon-l set type ping set interval 10 set timeout 2

set retry 3 end

2. Use the following command to add the virtual server that accepts HTTP sessions on port 8080 at the wan1 interface and load balances the traffic to three real servers. config firewall vip

Basic load balancing configuration example

edit Vserver-HTTP-1 set type server-load-balance set server-type http set ldb-method round-robin set extip 172.20.120.30 set extintf wan1 set extport 8080 set persistence http-cookie set monitor tcp-mon-1 config realservers edit 1 set ip 10.31.101.30

set port 80 next edit 2 set ip 10.31.101.40

set port 80 end edit 3 set ip 10.31.101.50

set port 80 end

end

3. Use the following command to add a security policy that includes the load balance virtual server as the destination address.

config firewall policy edit 0 set srcintf wan1 set srcaddr all set dstintf internal set dstaddr Vserver-HTTP-1 set action accept set schedule always set service ALL set nat enable set utm-status enable

set profile-protocol-options default set av-profile scan end