Iptables command
From DD-WRT Wiki
Revision as of 15:09, 1 December 2008 (edit) Chochem (Talk | contribs) ← Previous diff |
Revision as of 22:24, 4 January 2009 (edit) (undo) Soulstace (Talk | contribs) m (tables, chains, and targets) Next diff → |
||
Line 87: | Line 87: | ||
<br> | <br> | ||
Note: ppp0 is the WAN interface when PPPoE is used. This information is from [[IPv6]] page and quoted here: "The detailled configuration steps are targeted toward users with a basic DHCP connection for the WAN part. So, if using PPPoE will require replacing vlan1 with ppp0 in each instance. Other connection types will vary." | Note: ppp0 is the WAN interface when PPPoE is used. This information is from [[IPv6]] page and quoted here: "The detailled configuration steps are targeted toward users with a basic DHCP connection for the WAN part. So, if using PPPoE will require replacing vlan1 with ppp0 in each instance. Other connection types will vary." | ||
+ | |||
+ | ==Tables, Chains, and Targets== | ||
+ | ====Tables==== | ||
+ | The main tables we are concerned with are the "filter" table and the "nat" table. To list the contents of either table, do | ||
+ | |||
+ | iptables -t filter -L | ||
+ | iptables -t nat -L | ||
+ | |||
+ | The filter table is default and this includes chains like INPUT, OUTPUT, and FORWARD. The nat table is for Network Address Translation and it includes the PREROUTING and POSTROUTING chains. | ||
+ | ====Chains==== | ||
+ | INPUT is for packets destined to or entering the router's local sockets. | ||
+ | |||
+ | OUTPUT is for packets sourced from or leaving the router's local sockets. | ||
+ | |||
+ | FORWARD is for packets being forwarded through the router (e.g. packets not necessarily destined for local sockets). | ||
+ | |||
+ | PREROUTING is for manipulating packets before they are routed. | ||
+ | |||
+ | POSTROUTING is for manipulating packets after they are routed. | ||
+ | |||
+ | ====Targets==== | ||
+ | ACCEPT - packets are accepted/allowed | ||
+ | |||
+ | DROP - packets are dropped/denied | ||
+ | |||
+ | REJECT - packets are rejected/denied | ||
+ | |||
+ | logaccept - accepted packets are logged to /tmp/var/log/messages | ||
+ | |||
+ | logdrop - dropped packets are logged to /tmp/var/log/messages | ||
+ | |||
+ | DNAT is for altering packet's destination address. | ||
+ | |||
+ | SNAT is for altering packet's source address. | ||
==Examples== | ==Examples== | ||
Line 117: | Line 151: | ||
up for this purpose. | up for this purpose. | ||
- | ''Note: Simply adding a rule to the INPUT chain may | + | ''Note: Simply adding a rule to the INPUT chain may be enough to allow remote SSH access from the WAN. However, if your router is still in NAT/Gateway mode and wish to remap the SSH port to something different on the WAN side (say port 2222), you may Insert a PREROUTING rule instead. This is actually how the GUI does it when you enable remote WAN SSH management.'' |
- | iptables - | + | iptables -I INPUT -p tcp -d `nvram get lan_ipaddr` --dport 22 -j logaccept |
- | iptables -t nat -I PREROUTING -p tcp -m tcp --dport | + | iptables -t nat -I PREROUTING -p tcp -m tcp -d `nvram get wan_ipaddr` --dport 2222 -j DNAT --to-destination `nvram get lan_ipaddr`:80 |
Now if I type | Now if I type |
Revision as of 22:24, 4 January 2009
Iptables is a powerful administration tool for IPv4 packet filtering and NAT. It is used to set up, maintain, and inspect the tables of IP packet filter rules in the Linux kernel.
Iptables commands can be entered by command line interface, and/or saved as a Firewall script in the dd-wrt Administration panel. I tend to recommend testing and confirming your rules at the command line first. This way, if you happen to make a big mistake (like blocking access to the router), simply rebooting the router should repair it rather than having to do a hard reset. To get your rules to survive a reboot of the router, save them in a Firewall script as mentioned earlier.
I think we should have something about Firewall Builder on this page, since they're kind of related...
Basic Usage
iptables -[AD] chain rule-specification [options] iptables -[RI] chain rulenum rule-specification [options] iptables -D chain rulenum [options] iptables -[LFZ] [chain] [options] iptables -[NX] chain iptables -E old-chain-name new-chain-name iptables -P chain target [options] iptables -h (print this help information)
Commands
--append -A chain Append to chain --delete -D chain Delete matching rule from chain --delete -D chain rulenum Delete rule rulenum (1 = first) from chain --insert -I chain [rulenum] Insert in chain as rulenum (default 1=first) --replace -R chain rulenum Replace rule rulenum (1 = first) in chain --list -L [chain] List the rules in a chain or all chains --flush -F [chain] Delete all rules in chain or all chains --zero -Z [chain] Zero counters in chain or all chains --new -N chain Create a new user-defined chain --delete-chain -X [chain] Delete a user-defined chain --policy -P chain target Change policy on chain to target --rename-chain -E old-chain new-chain Change chain name, (moving any references)
Options
--proto -p [!] proto protocol: by number or name, eg. `tcp' --source -s [!] address[/mask] source specification --destination -d [!] address[/mask] destination specification --sport [!] port[:endport] source port (use `:' when specifying range) --dport [!] port[:endport] destination port --in-interface -i [!] input name[+] network interface name ([+] for wildcard) --jump -j target target for rule (may load target extension) --match -m match extended match (may load extension) --state state connection states to match: INVALID NEW ESTABLISHED RELATED --tcp-flags [!] mask match when the TCP flags are as specified: SYN ACK FIN RST URG PSH ALL NONE --numeric -n numeric output of addresses and ports --out-interface -o [!] output name[+] network interface name ([+] for wildcard) --table -t table table to manipulate (default: `filter') --verbose -v verbose mode --line-numbers print line numbers when listing --exact -x expand numbers (display exact values) --fragment -f match second or further fragments only --modprobe=<command> try to insert modules using this command --set-counters PKTS BYTES set the counter during insert/append --version -V print package version MAC v1.3.7 options: --mac-source [!] XX:XX:XX:XX:XX:XX Match source MAC address
Interfaces
When using the -i or -o to define the physical interfaces, remember that by default:
vlan0 is the 4 LAN ports
vlan1 is the WAN port (ppp0 is the WAN interface when PPPoE is used)
eth1 is the WIFI
br0 is a bridge connecting the 4 LAN and the WIFI together
Note: ppp0 is the WAN interface when PPPoE is used. This information is from IPv6 page and quoted here: "The detailled configuration steps are targeted toward users with a basic DHCP connection for the WAN part. So, if using PPPoE will require replacing vlan1 with ppp0 in each instance. Other connection types will vary."
Tables, Chains, and Targets
Tables
The main tables we are concerned with are the "filter" table and the "nat" table. To list the contents of either table, do
iptables -t filter -L iptables -t nat -L
The filter table is default and this includes chains like INPUT, OUTPUT, and FORWARD. The nat table is for Network Address Translation and it includes the PREROUTING and POSTROUTING chains.
Chains
INPUT is for packets destined to or entering the router's local sockets.
OUTPUT is for packets sourced from or leaving the router's local sockets.
FORWARD is for packets being forwarded through the router (e.g. packets not necessarily destined for local sockets).
PREROUTING is for manipulating packets before they are routed.
POSTROUTING is for manipulating packets after they are routed.
Targets
ACCEPT - packets are accepted/allowed
DROP - packets are dropped/denied
REJECT - packets are rejected/denied
logaccept - accepted packets are logged to /tmp/var/log/messages
logdrop - dropped packets are logged to /tmp/var/log/messages
DNAT is for altering packet's destination address.
SNAT is for altering packet's source address.
Examples
I think examples are the best way to demonstrate the use of iptables. (Take note, chains are to be typed in caps as shown!)
Listing the rules in a chain
First I want to view the rules on my INPUT chain, this is the first chain traffic coming into my router will hit.
iptables -L INPUT
You will find that it is really slow to list all the rules after you enter the above iptables command since it is doing reverse DNS loopups on IP addresses. You have to add -n option, in order to avoid long reverse DNS lookups.
iptables -L INPUT -n
To get a little more detailed list, with actual IP numbers and packet data, I can do
iptables -L INPUT -nv
Then I might want to add a rule so that I can ssh into my router from a specific host/address outside. So I might type the following:
iptables -A INPUT -p tcp -s 150.100.whatever.something --dport 22 -j logaccept
So I am saying: Append to the INPUT chain a rule allowing protocol tcp, with a source of <my external IP that i want access from> traffic destined for port 22 on my router, jump to logaccept. I could have used -j ACCEPT which simply jumps to ACCEPT, but in this case I want to log it just to keep track so I use logaccept, which is a chain we have set up for this purpose.
Note: Simply adding a rule to the INPUT chain may be enough to allow remote SSH access from the WAN. However, if your router is still in NAT/Gateway mode and wish to remap the SSH port to something different on the WAN side (say port 2222), you may Insert a PREROUTING rule instead. This is actually how the GUI does it when you enable remote WAN SSH management.
iptables -I INPUT -p tcp -d `nvram get lan_ipaddr` --dport 22 -j logaccept iptables -t nat -I PREROUTING -p tcp -m tcp -d `nvram get wan_ipaddr` --dport 2222 -j DNAT --to-destination `nvram get lan_ipaddr`:80
Now if I type
iptables -L INPUT
I see my shiny new rule appended to the INPUT chain. However, this is no good because in my case I have a rule blocking this traffic which occurs BEFORE the rule allowing it.
How do I change it? Simple.
First let's delete the rule we just made
iptables -L INPUT --line-numbers
will list the rules with their rule numbers. Let's say our rule is number 11
iptables -D INPUT 11
Clearly this Deletes rule number 11 from the input chain.
Now instead of Appending I am going to Insert my rule into the number 1 position.
iptables -I INPUT -p tcp -s 150.100.whatever.something --dport 22 -j logaccept
So now rule number 1 is my new rule and the other rules have all shifted down a position.
If I wanted to change the IP address or any other aspect of my ssh rule I could use the -R (Replace) option and simply type in the new rule, ie
iptables -R INPUT 1 -p tcp -s 100.100.200.100 --dport 22 -j ACCEPT
This would replace rule number 1 on the INPUT chain with the new rule which has a new source IP address and jumps to ACCEPT instead of logaccept.
One more example: I want to run a mini web server on my router. Let's assume that it is already running on port 8000 and I can access it from the LAN side, but not from the WAN side. With
iptables -I INPUT -p tcp -d 192.168.1.1 --dport 8000 -j logaccept
the port 8000 will be opened. But I also have to setup NAT PREROUTING, so that the kernel forwards all packets on port 8000 from the outside to itself, 192.168.1.1:
iptables -t nat -I PREROUTING -p tcp -d $(nvram get wan_ipaddr) --dport 8000 -j DNAT --to 192.168.1.1:8000
Port Forwarding to a specific LAN IP
Port Forwarding can be accomplished from within the web interface here. However, the very same thing can be done a bit differently (tested and working), via command line. --u3gyxap: Example with port 443 and IP 192.168.1.2
iptables -t nat -I PREROUTING -p tcp -d $(nvram get wan_ipaddr) --dport 443 -j DNAT --to 192.168.1.2:443 iptables -I FORWARD -p tcp -d 192.168.1.2 --dport 443 -j ACCEPT
If you want to restrict the source IP (a question that is asked a lot on the forums), add -s 123.45.67.89 to one of your rules (replacing the IP address with the real one of course).
iptables -t nat -I PREROUTING -p tcp -s 123.45.67.89 -d $(nvram get wan_ipaddr) --dport 443 -j DNAT --to 192.168.1.2:443 iptables -I FORWARD -p tcp -d 192.168.1.2 --dport 443 -j ACCEPT
This should make it so only one IP address is able to access your forwarded port from the Internet.
Deny access to a specific IP address
iptables -I FORWARD -d 123.123.123.123 -j DROP
Which would DROP all packets to the given IP. Useful to block access to what not. If you want to log the entry when the IP is blocked you would set DROP to logdrop.
Deny access to a specific Outbound IP address with logging
iptables -I OUTPUT -d 239.255.255.250 -j logdrop
This becomes useful if there is a program that wants to gain an outbound connection to a specific address, but you don't want to allow the connection. In this specific example Windows uses this IP incorrectly as a broadcast address (search Google for more info). While viewing your router logs you will see Windows broadcast to this IP several times per minute. By default the router passes the broadcast and announces to everyone outside of your router that your PC exists. This rule will block this specific outbound IP and add an entry into the router log.
Block SMTP traffic except to specified hosts
Simple Mail Transfer Protocol operates on tcp port 25.
/usr/sbin/iptables -I FORWARD 1 -p tcp -d safe.server1.com --dport 25 -j logaccept /usr/sbin/iptables -I FORWARD 2 -p tcp -d safe.server2.com --dport 25 -j logaccept /usr/sbin/iptables -I FORWARD 3 -p tcp --dport 25 -j logdrop
Which would accept and log all smtp traffic to safe.server1.com and safe.server2.com, while blocking and dropping all other outgoing smtp traffic.
Block outgoing SMTP traffic except from specified hosts
Simple Mail Transfer Protocol operates on tcp port 25.
iptables -I FORWARD 1 -p tcp -s 192.168.1.2 --dport 25 -j ACCEPT iptables -I FORWARD 2 -p tcp -s 192.168.1.1/24 --dport 25 -j REJECT
Which would accept outgoing SMTP traffic from your internal SMTP server (192.168.1.2) but reject outgoing SMTP traffic from all other hosts on your LAN (192.168.1.1/24). Useful to enforce all your LAN clients to use your internal SMTP server, as well as to block any viruses and spam-generating trojans from sending mail to remote servers on their own.
Change "REJECT" to "logdrop" or "ACCEPT" to "logaccept" to add logging.
Caution! This will also block internal users from using your external IP as their SMTP server.
Allow HTTP traffic only to specific domain(s)
Similarly, we can use the above method to filter other ports and protocols as well, such as standard web traffic operating on tcp port 80.
iptables -I FORWARD 1 -p tcp -d dd-wrt.com --dport 80 -j ACCEPT iptables -I FORWARD 2 -p tcp --dport 80 -j DROP
Which would accept all http traffic to dd-wrt.com, while blocking outgoing http traffic to anywhere else. If you wish to allow multiple sites, insert additional rules before the DROP (making sure to order and number them correctly).
Block all traffic except HTTP HTTPS and FTP
This example blocks everything except our normal web traffic, encrypted (ssl), and the file transfer protocol.
iptables -I FORWARD 1 -p tcp -m multiport --dport 21,80,443 -j ACCEPT iptables -I FORWARD 2 -m state --state ESTABLISHED,RELATED -j ACCEPT iptables -I FORWARD 3 -j DROP
Caution! Users are still able to get through the firewall if they are sly enough to change port numbers of their P2P or other application. In that case, you should consider using Access Restrictions to mitigate the possibility of that happening.
Restrict access by MAC address
In this example, we will demonstrate how to restrict access to the router's web interface by MAC address. In other words, only the computer having the specified MAC address should be able to access the web interface from the LAN.
First, if there are no Access Restrictions policies enabled and filtering by MAC addresses, you may need to insert the iptables mac module manually:
insmod ipt_mac iptables -I INPUT -p tcp --dport 80 -m mac ! --mac-source 00:12:34:56:78:9A -j REJECT --reject-with tcp-reset
Notice the ! (bang) which is another new concept introduced here. It means "NOT". So, by inspecting the rule closely, we see that it will REJECT packets destined to port 80 of the router so long as they do NOT originate from our computer with the desired MAC address.
Caution! As usual when dealing with MAC addresses, be aware that it is possible for malicious user(s) to spoof their MAC address with that of a trusted machine. You can help combat this by use of static ARP entries, VLANs, etc.
Modifying the TTL
The Time To Live is the maximum number of routers a packet will travel through before it is discarded. In certain situations, it may prove useful to modify it in order to make your network more reliable.
- Example 1: Set the incoming TTL to 10, before the router routes it into the LAN
iptables -t mangle -I PREROUTING -i vlan1 -j TTL --ttl-set 10
- Example 2: Set the outgoing TTL to 128, just as if a Windows machine was connected directly to the modem.
iptables -t mangle -I POSTROUTING -o vlan1 -j TTL --ttl-set 128
- Example 3: Try to hide the fact that an outgoing packet was routed, by incrementing the TTL by one.
iptables -t mangle -I POSTROUTING -o vlan1 -j TTL --ttl-inc 1
Caution
Adding iptables commands to your startup routine risks locking yourself out of the box with no option but to start over. If you are experimenting with new commands, you can insure yourself against this scenario by inserting a sleep command before the iptables command(s). This way you can grant yourself, say five minutes (sleep 300), before your commands take effect. If your commands do backfire and you are unable to log in to your box, simply restart it by switching it off and on again and you will have five minutes to get in.
See Also
One-to-one NAT
Port Blocking
Preventing Brute Force Attacks
External Resources
linux_journal_video_tut
http://linux.die.net/man/8/iptables
http://www.iptables.org/documentation/HOWTO//packet-filtering-HOWTO-7.html
http://www.iptables.org/documentation/HOWTO//netfilter-hacking-HOWTO.html