- Filtering by IP protocol; source and destination IP address; source and destination port for TCP and UDP traffic
- Limiting simultaneous connections on a per-rule basis
- Using p0f, an advanced passive OS/network fingerprinting utility to allow you to filter by operating system initiating the connection. For example, if you want to allow FreeBSD and Linux machines to the Internet, but block Windows machines, it is possible with the VirtualPF software, which passively detects the operating system in use.
- Option to log or not log traffic matching each rule.
- Highly flexible routing policy through gateway selection on a per-rule basis (for load balancing, failover, multiple WAN, etc.)
- Grouping and naming of IP addresses, networks and ports through the use of aliases. This helps keep the firewall ruleset clean and easy to understand, especially in environments with multiple public IPs and numerous servers.
- Transparent Layer 2 firewalling capability that can bridge interfaces and filter traffic between them, even allowing for an IP-less firewall (although an IP address may probably be desirable for management purposes).
- Packet normalization through scrubbing. The scrub documentation describes scrubbing as "the normalization of packets so there are no ambiguities in interpretation by the ultimate destination of the packet. The scrub directive also reassembles fragmented packets, protecting some operating systems from some forms of attack, and drops TCP packets that have invalid flag combinations."
- This feature is enabled in the VirtualPF software by default
- It can be disabled if necessary. (This option causes problems for some NFS implementations, but is otherwise safe and should be left enabled on most installations.)
- Option to turn off the firewall filter entirely if VirtualPF is to be used as a pure router.
VirtualPF is a stateful firewall
, and by default all rules are stateful. Because of this, information on open network connections is maintained in the firewall's state table. While most firewalls lack the ability to finely control the state table, VirtualPF has, in contrast, numerous features that allow granular control of the state table. Thanks to the following capabilities of OpenBSD's pf:
- Adjustable state table size – there are multiple production VirtualPF installations using several hundred thousand states. The default state table size varies according to the RAM installed in the system, but it can be increased on the fly to the desired size. Since each state takes approximately 1 KB of RAM, it is prudent to keep track of memory usage when sizing the state table. It should not be set it arbitrarily high.
- On a per-rule basis:
- Limit simultaneous client connections
- Limit states per host
- Limit new connections per second
- Define state timeout
- Define state type
- State types - the VirtualPF software offers multiple options for state handling.
- Keep state - Works with all protocols. Default for all rules.
- Sloppy state - Works with all protocols. Less strict state tracking, useful in cases of asymmetric routing.
- Modulate state - Works only with TCP. The VirtualPF software will generate strong Initial Sequence Numbers (ISNs) on behalf of the host.
- Synproxy state - Proxies incoming TCP connections to help protect servers from spoofed TCP SYN floods. This option includes the functionality of keep state and modulate state combined.
- None - Do not keep any state entries for this traffic. This is very rarely desirable, but is available because it can be useful under some limited circumstances.
- State table optimization options - pf offers four options for state table optimization.
- Normal - the default algorithm
- High latency - Useful for high latency links, such as satellite connections. Expires idle connections later than normal.
- Aggressive - Expires idle connections more quickly. More efficient use of hardware resources, but can drop legitimate connections.
- Conservative - Tries to avoid dropping legitimate connections at the expense of increased memory usage and CPU utilization.
Network Address Translation (NAT)
- Port forwards including ranges and the use of multiple public IPs
- 1:1 NAT for individual IPs or entire subnets.
- Outbound NAT
- Default settings NAT all outbound traffic to the WAN IP. In multiple WAN scenarios, the default settings NAT outbound traffic to the IP of the WAN interface being used.
- Advanced Outbound NAT allows this default behavior to be disabled, and enables the creation of very flexible NAT (or no NAT) rules.
- NAT Reflection - NAT reflection is possible so services can be accessed by public IP from internal networks.
PPTP / GRE Limitation - The state tracking code in pf for the GRE protocol can only track a single session per public IP per external server. This means if you use PPTP VPN connections, only one internal machine can connect simultaneously to a PPTP server on the Internet. A thousand machines can connect simultaneously to a thousand different PPTP servers, but only one simultaneously to a single server. The only available work around is to use multiple public IPs on your firewall, one per client, or to use multiple public IPs on the external PPTP server. This is not a problem with other types of VPN connections. PPTP is insecure and should no longer be used.
from OpenBSD allows for hardware failover. Two or more firewalls can be configured as a failover group. If one interface fails on the primary or the primary goes offline entirely, the secondary becomes active. The VirtualPF software also includes configuration synchronization capabilities, so you make your configuration changes on the primary and they automatically synchronize to the secondary firewall. pfsync
ensures the firewall's state table is replicated to all failover configured firewalls. This means your existing connections will be maintained in the case of failure, which is important to prevent network disruptions.
Only works with static public IPs, does not work with stateful failover using DHCP, PPPoE, or PPTP type WANs.
Multi-WAN functionality enables the use of multiple Internet connections, with load balancing and/or failover, for improved Internet availability and bandwidth usage distribution.
Server Load Balancing
Server load balancing is used to distribute load between multiple servers. This is commonly used with web servers, mail servers, and others. Servers that fail to respond to ping requests or TCP port connections are removed from the pool.
Virtual Private Network (VPN)
The VirtualPF software offers three options for VPN connectivity, IPsec
, and PPTP
IPsec allows connectivity with any device supporting standard IPsec. This is most commonly used for site to site connectivity to other VirtualPF installations, other open source firewalls (m0n0wall, etc.), and most all commercial firewall solutions (Cisco, Juniper, etc.). It can also be used for mobile client connectivity.
OpenVPN is a flexible, powerful SSL VPN solution supporting a wide range of client operating systems. See theOpenVPN website
for details on its abilities.
PPTP was a popular VPN option because nearly every OS has a built in PPTP client, including every Windows release since Windows 95 OSR2. However, it's now considered insecure and should not be used. See this Wikipedia article
for more information on the PPTP protocol.
Because of limitations in pf NAT, when the PPTP Server is enabled, PPTP clients cannot use the same public IP for outbound PPTP connections. This means if you have only one public IP, and use the PPTP Server, PPTP clients inside your network will not work. The work around is to use a second public IP with Advanced Outbound NAT for your internal clients. See also the PPTP limitation under NAT on this page.
The VirtualPF software offers a PPPoE server. For more information on the PPPoE protocol, see this Wikipedia entry
. A local user database can be used for authentication, and RADIUS authentication with optional accounting is also supported.
Reporting and Monitoring
The RRD graphs in the VirtualPF software maintain historical information on the following.
- CPU utilization
- Total throughput
- Firewall states
- Individual throughput for all interfaces
- Packets per second rates for all interfaces
- WAN interface gateway(s) ping response times
- Traffic shaper queues on systems with traffic shaping enabled
Real Time Information
Historical information is important, but sometimes it's more important to see real time information.
- SVG graphs are available that show real time throughput for each interface.
- For traffic shaper users, the Status -> Queues screen provides a real time display of queue usage using AJAX updated gauges.
- The front page includes AJAX gauges for display of real time CPU, memory, swap and disk usage, and state table size.
A Dynamic DNS client is included to allow you to register your public IP with a number of dynamic DNS service providers.
- Custom - allowing defining update method for providers not specifically listed here.
- Route 53
A client is also available for RFC 2136 dynamic DNS updates, for use with DNS servers like BIND which support this means of updating.
Captive portal allows you to force authentication, or redirection to a click through page for network access. This is commonly used on hot spot networks, but is also widely used in corporate networks for an additional layer of security on wireless or Internet access. For more information on captive portal technology in general, see theWikipedia article
on the topic. The following is a list of features in the VirtualPF Captive Portal.
- Maximum concurrent connections - Limit the number of connections to the portal itself per client IP. This feature prevents a denial of service from client PCs sending network traffic repeatedly without authenticating or clicking through the splash page.
- Idle timeout - Disconnect clients who are idle for more than the defined number of minutes.
- Hard timeout - Force a disconnect of all clients after the defined number of minutes.
- Logon pop up window - Option to pop up a window with a log off button.
- URL Redirection - after authenticating or clicking through the captive portal, users can be forcefully redirected to the defined URL.
- MAC filtering - by default, VirtualPF filters using MAC addresses. If you have a subnet behind a router on a captive portal enabled interface, every machine behind the router will be authorized after one user is authorized. MAC filtering can be disabled for these scenarios.
- Authentication options - There are three authentication options available.
- No authentication - This means the user just clicks through your portal page without entering credentials.
- Local user manager - A local user database can be configured and used for authentication.
- RADIUS authentication - This is the preferred authentication method for corporate environments and ISPs. It can be used to authenticate from Microsoft Active Directory and numerous other RADIUS servers.
- RADIUS capabilities
- Forced re-authentication
- Able to send Accounting updates
- RADIUS MAC authentication allows captive portal to authenticate to a RADIUS server using the client's MAC address as the user name and password.
- Allows configuration of redundant RADIUS servers.
- HTTP or HTTPS - The portal page can be configured to use either HTTP or HTTPS.
- Pass-through MAC and IP addresses - MAC and IP addresses can be white listed to bypass the portal. Any machines with NAT port forwards will need to be bypassed so the reply traffic does not hit the portal. You may wish to exclude some machines for other reasons.
- File Manager - This allows you to upload images for use in your portal pages.
"Reverse" portal, i.e. capturing traffic originating from the Internet and entering your network, is not possible. Only entire IP and MAC addresses can be excluded from the portal, not individual protocols and ports.
DHCP Server and Relay
The VirtualPF software includes both DHCP Server and Relay functionality
Original source: https://www.pfsense.org/about-pfsense/features.html