IPtables DDoS Protection: The Best Rules to Mitigate DDoS Attacks
Written by BiRU Friday, 10 February 2017 09:38
There are different ways of building your own anti-DDoS rules for iptables. We will be discussing the most effective iptables DDoS protection methods in this comprehensive tutorial.
This guide will teach you how to:
- Select the best iptables table and chain to stop DDoS attacks
- Tweak your kernel settings to mitigate the effects of DDoS attacks
- Use iptables to block most TCP-based DDoS attacks
- Use iptables SYNPROXY to block SYN floods
Please note that this article is written for professionals who deal with Linux servers on a daily basis. If you just want to protect your online application from DDoS attacks, you can use our remote protection, a VPS with DDoS protection or a DDoS protected bare metal server.
While one can do a lot with iptables to block DDoS attacks, there isn't a way around actual hardware firewalls (we recently reviewed RioRey DDoS mitigation hardware) to detect and stop large DDoS floods. However, it isn't impossible to filter most bad traffic at line rate using iptables.
JavaPipe provides advanced DDoS protection solutions against all types of DDoS attacks. View DDoS Mitigation Solutions
We'll only cover protection from TCP-based attacks. Most UDP-based attacks are amplified reflection attacks that will exhaust the network interface card of any common server. The only mitigation approach that makes sense against these types of attacks is to block them at the edge or core network or even at the carrier already. If they are able to reach your server, there isn't much you can do against those multi-Gbit/s attacks except to move to a DDoS protected network.
What Is IPtables?
netfilter iptables (soon to be replaced by nftables) is a user-space command line utility to configure kernel packet filtering rules developed by netfilter. It's the default firewall management utility on Linux systems - everyone working with Linux systems should be familiar with it or have at least heard of it.
iptables can be used to filter certain packets, block source or destination ports and IP addresses, forward packets via NAT and a lot of other things. Most commonly it's used to block destination ports and source IP addresses.
Why Your IPtables Anti-DDoS Rules Suck
To understand why your current iptables rules to prevent DDoS attacks suck, we first have to dig into how iptables works.
iptables is a command line tool used to set up and control the tables of IP packet filter rules. There are different tables for different purposes.
IPtables Tables
Filter: The filter table is the default and most commonly used table that rules go to if you don't use the -t (--table) option.
Nat: This table is used for Network Address Translation (NAT). If a packet creates a new connection, the nat table gets checked for rules.
Mangle: The mangle table is used to modify or mark packets and their header information.
Raw: This table's purpose is mainly to exclude certain packets from connection tracking using the NOTRACK target.
As you can see there are four different tables on an average Linux system that doesn't have non-standard kernel modules loaded. Each of these tables supports a different set of iptables chains.
IPtables Chains
PREROUTING: raw, nat, mangle
- Applies to packets that enter the network interface card (NIC)
INPUT: filter, mangle
- Applies to packets destined to a local socket
FORWARD: filter, mangle
- Applies to packets that are being routed through the server
OUTPUT: raw, filter, nat, mangle
- Applies to packets that the server sends (locally generated)
POSTROUTING: nat, mangle
- Applies to packets that leave the server
Depending on what kind of packets you want to block or modify, you select a certain iptables table and a chain that the selected table supports.
Of course, we're still missing an explanation of iptables targets (ACCEPT, DROP, REJECT, etc.), but we're assuming that if you're reading this article, you already know how to deal with iptables.
We're going to explain why your iptables rules suck to stop DDoS and not teach you how to use iptables. Let's get back to that.
If you want to block a DDoS attack with iptables, performance of the iptables rules is extremely important. Most TCP-based DDoS attack types use a high packet rate, meaning the sheer number of packets per second is what causes the server to go down. That's why you want to make sure that you can process and block as many packets per second as possible.
You'll find that most if not all guides on how to block DDoS attacks using iptables use the filter table and the INPUT chain for anti-DDoS rules. The issue with this approach is that the INPUT chain is only processed after the PREROUTING and FORWARD chains and therefore only applies if the packet doesn't match any of these two chains.
This causes a delay in the filtering of the packet which consumes resources. In conclusion, to make our rules as effective as possible, we need to move our anti-DDoS rules as far up the chains as possible. The first chain that can apply to a packet is the PREROUTING chain, so ideally we'll want to filter the bad packets in this chain already.
However, the filter table doesn't support the PREROUTING chain. To get around this problem, we can simply use the mangle table instead of the filter table for our anti-DDoS iptables rules. It supports most if not all rules that the filter table supports while also supporting all iptables chains.
So you want to know why your iptables DDoS protection rules suck? It's because you use the filter table and the INPUT chain to block the bad packets! The best solution to dramatically increase the performance of your iptables rules and therefore the amount of (TCP) DDoS attack traffic they can filter is to use the mangle table and the PREROUTING chain!
The Best Linux Kernel Settings to Mitigate DDoS
Another common mistake is that people don't use optimized kernel settings to better mitigate the effects of DDoS attacks. Note that this guide focuses on CentOS 7 as the operating system of choice. CentOS 7 includes a recent version of iptables and support of the new SYNPROXY target.
We won't cover every single kernel setting that you need to adjust in order to better mitigate DDoS with iptables. Instead, we provide a set of CentOS 7 kernel settings that we would use. Just put the below in your /etc/sysctl.conf file and apply the settings with sysctl -p.
Anti-DDoS Kernel Settings (sysctl.conf)
kernel.printk = 4 4 1 7 kernel.panic = 10 kernel.sysrq = 0 kernel.shmmax = 4294967296 kernel.shmall = 4194304 kernel.core_uses_pid = 1 kernel.msgmnb = 65536 kernel.msgmax = 65536 vm.swappiness = 20 vm.dirty_ratio = 80 vm.dirty_background_ratio = 5 fs.file-max = 2097152 net.core.netdev_max_backlog = 262144 net.core.rmem_default = 31457280 net.core.rmem_max = 67108864 net.core.wmem_default = 31457280 net.core.wmem_max = 67108864 net.core.somaxconn = 65535 net.core.optmem_max = 25165824 net.ipv4.neigh.default.gc_thresh1 = 4096 net.ipv4.neigh.default.gc_thresh2 = 8192 net.ipv4.neigh.default.gc_thresh3 = 16384 net.ipv4.neigh.default.gc_interval = 5 net.ipv4.neigh.default.gc_stale_time = 120 net.netfilter.nf_conntrack_max = 10000000 net.netfilter.nf_conntrack_tcp_loose = 0 net.netfilter.nf_conntrack_tcp_timeout_established = 1800 net.netfilter.nf_conntrack_tcp_timeout_close = 10 net.netfilter.nf_conntrack_tcp_timeout_close_wait = 10 net.netfilter.nf_conntrack_tcp_timeout_fin_wait = 20 net.netfilter.nf_conntrack_tcp_timeout_last_ack = 20 net.netfilter.nf_conntrack_tcp_timeout_syn_recv = 20 net.netfilter.nf_conntrack_tcp_timeout_syn_sent = 20 net.netfilter.nf_conntrack_tcp_timeout_time_wait = 10 net.ipv4.tcp_slow_start_after_idle = 0 net.ipv4.ip_local_port_range = 1024 65000 net.ipv4.ip_no_pmtu_disc = 1 net.ipv4.route.flush = 1 net.ipv4.route.max_size = 8048576 net.ipv4.icmp_echo_ignore_broadcasts = 1 net.ipv4.icmp_ignore_bogus_error_responses = 1 net.ipv4.tcp_congestion_control = htcp net.ipv4.tcp_mem = 65536 131072 262144 net.ipv4.udp_mem = 65536 131072 262144 net.ipv4.tcp_rmem = 4096 87380 33554432 net.ipv4.udp_rmem_min = 16384 net.ipv4.tcp_wmem = 4096 87380 33554432 net.ipv4.udp_wmem_min = 16384 net.ipv4.tcp_max_tw_buckets = 1440000 net.ipv4.tcp_tw_recycle = 0 net.ipv4.tcp_tw_reuse = 1 net.ipv4.tcp_max_orphans = 400000 net.ipv4.tcp_window_scaling = 1 net.ipv4.tcp_rfc1337 = 1 net.ipv4.tcp_syncookies = 1 net.ipv4.tcp_synack_retries = 1 net.ipv4.tcp_syn_retries = 2 net.ipv4.tcp_max_syn_backlog = 16384 net.ipv4.tcp_timestamps = 1 net.ipv4.tcp_sack = 1 net.ipv4.tcp_fack = 1 net.ipv4.tcp_ecn = 2 net.ipv4.tcp_fin_timeout = 10 net.ipv4.tcp_keepalive_time = 600 net.ipv4.tcp_keepalive_intvl = 60 net.ipv4.tcp_keepalive_probes = 10 net.ipv4.tcp_no_metrics_save = 1 net.ipv4.ip_forward = 0 net.ipv4.conf.all.accept_redirects = 0 net.ipv4.conf.all.send_redirects = 0 net.ipv4.conf.all.accept_source_route = 0 net.ipv4.conf.all.rp_filter = 1
These sysctl.conf settings help to maximize the performance of your server under DDoS as well as the effectiveness of the iptables rules that we're going to provide in this guide.
The Actual IPtables Anti-DDoS Rules
Considering you now know that you need to use the mangle table and the PREROUTING chain as well as optimized kernel settings to mitigate the effects of DDoS attacks, we'll now move on to a couple of example rules to mitigate most TCP DDoS attacks.
DDoS attacks are complex. There are many different types of DDoS and it's close to impossible to maintain signature-based rules against all of them. But luckily there is something called connection tracking (nf_conntrack kernel module), which can help us to mitigate almost any TCP-based DDoS attack that doesn't use SYN packets that seem legitimate. This includes all types of ACK and SYN-ACK DDoS attacks as well as DDoS attacks that use bogus TCP flags.
We'll start with just five simple iptables rules that will already drop many TCP-based DDoS attacks.
Block Invalid Packets
iptables -t mangle -A PREROUTING -m conntrack --ctstate INVALID -j DROP
This rule blocks all packets that are not a SYN packet and don't belong to an established TCP connection.
Block New Packets That Are Not SYN
iptables -t mangle -A PREROUTING -p tcp ! --syn -m conntrack --ctstate NEW -j DROP
This blocks all packets that are new (don't belong to an established connection) and don't use the SYN flag. This rule is similar to the "Block Invalid Packets" one, but we found that it catches some packets that the other one doesn't.
Block Uncommon MSS Values
iptables -t mangle -A PREROUTING -p tcp -m conntrack --ctstate NEW -m tcpmss ! --mss 536:65535 -j DROP
The above iptables rule blocks new packets (only SYN packets can be new packets as per the two previous rules) that use a TCP MSS value that is not common. This helps to block dumb SYN floods.
Block Packets With Bogus TCP Flags
iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,SYN,RST,PSH,ACK,URG NONE -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,SYN FIN,SYN -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags SYN,RST SYN,RST -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags SYN,FIN SYN,FIN -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,RST FIN,RST -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,ACK FIN -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ACK,URG URG -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ACK,FIN FIN -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ACK,PSH PSH -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL ALL -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL NONE -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL FIN,PSH,URG -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL SYN,FIN,PSH,URG -j DROP iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL SYN,RST,ACK,FIN,URG -j DROP
The above ruleset blocks packets that use bogus TCP flags, ie. TCP flags that legitimate packets wouldn't use.
Block Packets From Private Subnets (Spoofing)
iptables -t mangle -A PREROUTING -s 224.0.0.0/3 -j DROP iptables -t mangle -A PREROUTING -s 169.254.0.0/16 -j DROP iptables -t mangle -A PREROUTING -s 172.16.0.0/12 -j DROP iptables -t mangle -A PREROUTING -s 192.0.2.0/24 -j DROP iptables -t mangle -A PREROUTING -s 192.168.0.0/16 -j DROP iptables -t mangle -A PREROUTING -s 10.0.0.0/8 -j DROP iptables -t mangle -A PREROUTING -s 0.0.0.0/8 -j DROP iptables -t mangle -A PREROUTING -s 240.0.0.0/5 -j DROP iptables -t mangle -A PREROUTING -s 127.0.0.0/8 ! -i lo -j DROP
These rules block spoofed packets originating from private (local) subnets. On your public network interface you usually don't want to receive packets from private source IPs. These rules assume that your loopback interface uses the 127.0.0.0/8 IP space.
These five sets of rules alone already block many TCP-based DDoS attacks at very high packet rates. With the kernel settings and rules mentioned above, you'll be able to filter ACK and SYN-ACK attacks at line rate.
Additional Rules
iptables -t mangle -A PREROUTING -p icmp -j DROP
This drops all ICMP packets. ICMP is only used to ping a host to find out if it's still alive. Because it's usually not needed and only represents another vulnerability that attackers can exploit, we block all ICMP packets to mitigate Ping of Death (ping flood), ICMP flood and ICMP fragmentation flood.
iptables -A INPUT -p tcp -m connlimit --connlimit-above 80 -j REJECT --reject-with tcp-reset
This iptables rule helps against connection attacks. It rejects connections from hosts that have more than 80 established connections. If you face any issues you should raise the limit as this could cause troubles with legitimate clients that establish a large number of TCP connections.
iptables -A INPUT -p tcp -m conntrack --ctstate NEW -m limit --limit 60/s --limit-burst 20 -j ACCEPT iptables -A INPUT -p tcp -m conntrack --ctstate NEW -j DROP
Limits the new TCP connections that a client can establish per second. This can be useful against connection attacks, but not so much against SYN floods because the usually use an endless amount of different spoofed source IPs.
iptables -t mangle -A PREROUTING -f -j DROP
This rule blocks fragmented packets. Normally you don't need those and blocking fragments will mitigate UDP fragmentation flood. But most of the time UDP fragmentation floods use a high amount of bandwidth that is likely to exhaust the capacity of your network card, which makes this rule optional and probably not the most useful one.
iptables -A INPUT -p tcp --tcp-flags RST RST -m limit --limit 2/s --limit-burst 2 -j ACCEPT iptables -A INPUT -p tcp --tcp-flags RST RST -j DROP
This limits incoming TCP RST packets to mitigate TCP RST floods. Effectiveness of this rule is questionable.
Mitigating SYN Floods With SYNPROXY
SYNPROXY is a new target of iptables that has been added in Linux kernel version 3.12 and iptables 1.4.21. CentOS 7 backported the feature and it's available in its 3.10 default kernel.
The purpose of SYNPROXY is to check whether the host that sent the SYN packet actually establishes a full TCP connection or just does nothing after it sent the SYN packet. If it does nothing, it discards the packet with minimal performance impact.
Did you know we can protect your whole network infrastructure by re-routing traffic through our PoP?
Contact us for a quote on BGP DDoS protection for networks!
While the iptables rules that we provided above already block most TCP-based attacks, the attack type that can still slip through them if sophisticated enough is a SYN flood. It's important to note that the performance of the rules will always be better if we find a certain pattern or signature to block, such as packet length (-m length), TOS (-m tos), TTL (-m ttl) or strings and hex values (-m string and -m u32 for the more advanced users). But in some rare cases that's not possible or at least not easy to achieve. So in these cases, you can make use of SYNPROXY.
Here are iptables SYNPROXY rules that help mitigate SYN floods that bypass our other rules:
iptables -t raw -A PREROUTING -p tcp -m tcp --syn -j CT --notrack iptables -A INPUT -p tcp -m tcp -m conntrack --ctstate INVALID,UNTRACKED -j SYNPROXY --sack-perm --timestamp --wscale 7 --mss 1460 iptables -A INPUT -m conntrack --ctstate INVALID -j DROP
These rules apply to all ports. If you want to use SYNPROXY only on certain TCP ports that are active (recommended - also you should block all TCP ports that are not in use using the mangle table and PREROUTING chain), you can just add --dport 80 to each of the rules if you want to use SYNPROXY on port 80 only.
To verify that SYNPROXY is working, you can do watch -n1 cat /proc/net/stat/synproxy. If the values change when you establish a new TCP connection to the port you use SYNPROXY on, it works.
The Complete IPtables Anti-DDoS Rules
If you don't want to copy & paste each single rule we discussed in this article, you can use the below ruleset for basic DDoS protection of your Linux server.
### 1: Drop invalid packets ### /sbin/iptables -t mangle -A PREROUTING -m conntrack --ctstate INVALID -j DROP ### 2: Drop TCP packets that are new and are not SYN ### /sbin/iptables -t mangle -A PREROUTING -p tcp ! --syn -m conntrack --ctstate NEW -j DROP ### 3: Drop SYN packets with suspicious MSS value ### /sbin/iptables -t mangle -A PREROUTING -p tcp -m conntrack --ctstate NEW -m tcpmss ! --mss 536:65535 -j DROP ### 4: Block packets with bogus TCP flags ### /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,SYN,RST,PSH,ACK,URG NONE -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,SYN FIN,SYN -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags SYN,RST SYN,RST -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags SYN,FIN SYN,FIN -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,RST FIN,RST -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags FIN,ACK FIN -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ACK,URG URG -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ACK,FIN FIN -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ACK,PSH PSH -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL ALL -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL NONE -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL FIN,PSH,URG -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL SYN,FIN,PSH,URG -j DROP /sbin/iptables -t mangle -A PREROUTING -p tcp --tcp-flags ALL SYN,RST,ACK,FIN,URG -j DROP ### 5: Block spoofed packets ### /sbin/iptables -t mangle -A PREROUTING -s 224.0.0.0/3 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 169.254.0.0/16 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 172.16.0.0/12 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 192.0.2.0/24 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 192.168.0.0/16 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 10.0.0.0/8 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 0.0.0.0/8 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 240.0.0.0/5 -j DROP /sbin/iptables -t mangle -A PREROUTING -s 127.0.0.0/8 ! -i lo -j DROP ### 6: Drop ICMP (you usually don't need this protocol) ### /sbin/iptables -t mangle -A PREROUTING -p icmp -j DROP ### 7: Drop fragments in all chains ### /sbin/iptables -t mangle -A PREROUTING -f -j DROP ### 8: Limit connections per source IP ### /sbin/iptables -A INPUT -p tcp -m connlimit --connlimit-above 111 -j REJECT --reject-with tcp-reset ### 9: Limit RST packets ### /sbin/iptables -A INPUT -p tcp --tcp-flags RST RST -m limit --limit 2/s --limit-burst 2 -j ACCEPT /sbin/iptables -A INPUT -p tcp --tcp-flags RST RST -j DROP ### 10: Limit new TCP connections per second per source IP ### /sbin/iptables -A INPUT -p tcp -m conntrack --ctstate NEW -m limit --limit 60/s --limit-burst 20 -j ACCEPT /sbin/iptables -A INPUT -p tcp -m conntrack --ctstate NEW -j DROP ### 11: Use SYNPROXY on all ports (disables connection limiting rule) ### #/sbin/iptables -t raw -A PREROUTING -p tcp -m tcp --syn -j CT --notrack #/sbin/iptables -A INPUT -p tcp -m tcp -m conntrack --ctstate INVALID,UNTRACKED -j SYNPROXY --sack-perm --timestamp --wscale 7 --mss 1460 #/sbin/iptables -A INPUT -m conntrack --ctstate INVALID -j DROP
Bonus Rules
Here are some more iptables rules that are useful to increase the overall security of a Linux server:
### SSH brute-force protection ### /sbin/iptables -A INPUT -p tcp --dport ssh -m conntrack --ctstate NEW -m recent --set /sbin/iptables -A INPUT -p tcp --dport ssh -m conntrack --ctstate NEW -m recent --update --seconds 60 --hitcount 10 -j DROP ### Protection against port scanning ### /sbin/iptables -N port-scanning /sbin/iptables -A port-scanning -p tcp --tcp-flags SYN,ACK,FIN,RST RST -m limit --limit 1/s --limit-burst 2 -j RETURN /sbin/iptables -A port-scanning -j DROP
Conclusion
This tutorial demonstrates some of the most powerful and effective methods to stop DDoS attacks using iptables. We've successfully mitigated DDoS attacks that peaked at multiple million packets per second using these iptables rules.
Every single guide on the same topic that we had researched provided inefficient methods to stop DDoS traffic or only a very limited number of iptables rules. If used correctly, iptables is an extremely powerful tool that's able to block different types of DDoS attacks at line-rate of 1GigE NICs and close to line-rate of 10GigE NICs. Don't underestimate the power of iptables!
