Having a great defense involves proper detection and recognition of an attack. In our security world we have great IDS tools to properly recognize when we are being attacked as well as firewalls to prevent such attacks from happening. However, certain attacks are not blindly thrown at you - a good attacker knows that a certain amount of reconnaissance and knowledge about your defenses greatly increases the chances of a successful attack. How would you know if someone is scanning your defenses? Is there any way to properly respond to such scans? You bet there is...
 Eckie S. |
The Port Scan Attack Detector (psad) is an excellent tool for detecting
various types of suspicious traffic, including port scans from popular tools
such as Nmap, DDoS attacks, and other efforts to brute force certain protocols
on your system. By analyzing firewall logs, psad can not only pick up on
certain attack patterns, but even manipulate firewall rules to properly respond
to suspicious activity.
This article will walk the reader through an EnGarde Secure Linux
implementation of psad, from the initial iptables rules setup to the deployment
of psad on the server side. By the end of the article, the user will be able to
detect certain Nmap scans and have psad respond to these scans by blocking the
source.
Prerequisites
+---------------------------------------------------------------------------You will need:
- A machine with EnGarde Secure Community 3.0.18 or above installed to do your
development on. These commands should NOT be run on a production server
since psad will eventually deny any type of access from the remote scanning
machine!
- A separate machine on the same network with Nmap installed on it. You will
be running certain scans on the server from this machine.
Once you have all the above you may log in as root, transition over to sysadm_r, and disable SELinux:
[psad_server]# newrole -r sysadm_r
Authenticating root.
Password:
[psad_server]# setenforce 0
Throughout the HowTo, the server will be referred to as psad_server and the
Nmap scanning machine as nmap_scanner.
Install psad
+---------------------------------------------------------------------------EnGarde Secure Linux makes the installation of psad a breeze due to its
Guardian Digital Secure Network (GDSN). You can install the package through
the command line:
[psad_server]# apt-get install psad
...or login to WebTool and download the package from the package manager
interface.
We shall get around to the setup of psad after we configure the firewalls on
psad_server to log packets:
iptables Rules Setup
+---------------------------------------------------------------------------Since iptables is installed out of the box on EnGarde Secure Linux, you only
have to run two simple commands to start logging packets with iptables:
[psad_server]# iptables -A INPUT -j LOG
[psad_server]# iptables -A FORWARD -j LOG
From here on out incoming packets (especially those of Nmap scans) will be
logged. Let's see if we can start detecting such scans by setting up psad to do
so.
psad Configuration
+---------------------------------------------------------------------------On psad_server, use your favorite editor to modify the /etc/psad/psad.conf
file. We're interested in the following tunables:
EMAIL_ADDRESSES
HOSTNAME
SYSLOG_DAEMON
ETC_SYSLOGNG_CONF
The EMAIL_ADDRESSES should be whichever email addresses you wish to have psad
send feedback to. This feedback includes error messages and alerts of potential
dangerous scans depending on danger levels which can be fine-tuned for your
purposes.
- The HOSTNAME tunable will be the hostname of the psad_server machine.
- The SYSLOG_DAEMON refers to the logging daemon for the machine. For EnGarde
Secure Linux, this should be set to 'syslog-ng'.
- The ETC_SYSLOGNG_CONF refers to the direct path of the syslog-ng daemon's
configuration file. For EnGarde Secure Linux, this should be set to
'/etc/syslog-ng.conf'.
- Once you've properly configured those tunables, you can start the psad daemon:
[psad_server]# /etc/init.d/psad start
[ SUCCESSFUL ] psad Daemons
Note: As far as danger levels are concerned, these range from one to five
and are assigned to the IP addresses from which an attack or scan is
detected. They are assigned based on the number of packets sent, port range, the
time interval of the scan, whether or not the signatures of the packets
match up with psad signature attacks, and the IP address where the packet
originated from. Depending on the number of such packets, a level is assigned as
per the configuration file. For more information on danger levels and ideas
for fine-tuning them, please refer to the resources at the end of the article.
psad - Active Detection
+---------------------------------------------------------------------------We will now use psad to detect certain Nmap scans. On the Nmap scanning
machine, run a TCP connect() scan by executing the following:
[nmap_scanner] nmap -sT 1.2.3.4
Replace 1.2.3.4 with the IP address of your psad_server.
If we check the /var/log/psad/fwdata file on the psad_server, you will find
the following:
Feb 2 11:58:11 psad_server kernel: IN=eth0 OUT MAC=00:0c:29:78:22:73:00:0c:76:4b:f6:3e:08:00 SRC=5.6.7.8
DST=1.2.3.4 LEN=60 TOS=0x00 PREC=0x00 TTL=64 ID=23609 DF PROTO=TCP
SPT=49021 DPT=113 WINDOW=5840 RES=0x00 SYN URGP=0
We can see that SRC will have the IP address of the nmap_scanner machine, and
DST will have the address of the psad_server. Also note that PROTO=TCP, showing
that the attack was a TCP connect() scan.
If you had previously configured psad to send email alerts, you will begin
receiving emails concerning this scan showing lots more data than these log
messages can ever produce. There are configuration tunables in the
/etc/psad/psad.conf file to limit and even disable email:
EMAIL_LIMIT
ALERTING_METHODS
EMAIL_ALERT_DANGER_LEVEL
EMAIL_LIMIT defines the maximum number of emails a configured user will
receive for a given IP address.
ALERTING_METHODS can be set to noemail, nosyslog, and ALL, depending on
whether you want only syslog-ng messages, email alerts, or both.
EMAIL_ALERT_DANGER_LEVEL is the minimum danger level that must be hit in order
for psad to send email alerts concerning a detection. The default setting is
one, so you can expect lots of emails for this tutorial's purpose.
Here is an example email showing psad output of the previous Nmap scan:
Subject: [psad-alert] DL2 src: nmap_scanner.yournetwork.com dst:
psad_server.yournetwork.com
Danger level: [2] (out of 5)
Scanned UDP ports: [32772: 1 packets, Nmap: -sU]
iptables chain: INPUT, 1 packets
Source: 5.6.7.8
DNS: nmap_scanner.yournetwork.com
OS guess: Linux (2.4.x kernel)
Destination: 1.2.3.4
DNS: psad_server.yournetwork.com
Overall scan start: Mon Feb 2 11:57:19 2008
Total email alerts: 2
Complete TCP range: [64-49400]
Complete UDP range: [32772]
Syslog hostname: unknown
Global stats: chain: interface: TCP: UDP: ICMP:
INPUT eth0 40 1 0
[+] TCP scan signatures:
"P2P Napster Client Data communication attempt"
dst port: 5555 (no server bound to local port)
flags: SYN
sid: 564
chain: INPUT
packets: 1
classtype: policy-violation
As you can see, psad does a wonderful job of taking packet data from logs, analyzing it and producing useful information on the type of scans used. psad - Active Defense +--------------------------------------------------------------------------- One of the more prominent features of psad is its active defense implementation - being able to detect Nmap scans is nice, but how do you respond? Let's configure psad to automatically block the source of such scans upon detection. Before implementing this feature, it is obvious for certain security veterans who are reading this article that there is a definite tradeoff for enforcing an active response policy. Although malicious traffic will be blocked, there is always the risk of blocking out valid traffic. Certain attackers can exploit active defenses and turn it against the target by attempting to spoof valid addresses, thus blocking out otherwise harmless traffic. This only happens in cases where the active response system has been configured to respond to nearly ALL types of potentially harmful traffic, including port scans or port sweeps. This also applies to traffic which does not require bidirectional communication with the target. A better strategy to employ is to only respond to traffic where bidirectional communication is required i.e. TCP connections. Even then, one must take care to tailor their active response to certain types of TCP connections, such as attempted SQL injection attacks, etc. Please be sure you are absolutely positive of how your detection scheme is working before deploying an active defense. Using your favorite editor, modify the /etc/psad/psad.conf file. We're interested in the following tunables: ENABLE_AUTO_IDS AUTO_IDS_DANGER_LEVEL ENABLE_AUTO_IDS should be set to 'Y' to enable the automated IDS response. AUTO_IDS_DANGER_LEVEL, for this HowTo's sake, will be set to '3'. This danger level is customizable and the setting we use in this HowTo is for demonstration purposes only. Restart the psad on the psad_server: [psad_server]# /etc/init.d/psad restart [ SUCCESSFUL ] psadwatchd Daemon [ SUCCESSFUL ] psad Daemon [ SUCCESSFUL ] kmsgsd Daemon [ SUCCESSFUL ] psad Daemons From the nmap_scanner machine, we'll run an Nmap SYN scan along with the '-P0' switch - this type of scan uses no ping and does not fully complete a TCP connection, resulting in fast scans. This usually requires root privileges, and is considered more of a dangerous scan - just the type of scan that psad detects at a higher danger level. [nmap_scanner]# nmap -sS -P0 -n 1.2.3.4 Replace the '1.2.3.4' with the IP address of your psad_server machine. psad will detect the SYN scans, and since the danger level of this scan is 3, it manipulates the iptables rules to block the source of the scans. This can be verified on the psad_server by running the following command: [psad_server]# psad --fw-list [+] Listing chains from IPT_AUTO_CHAIN keywords... Chain PSAD_BLOCK_INPUT (1 references) pkts bytes target prot opt in out source destination 820 36080 DROP all -- * * 5.6.7.8 0.0.0.0/0 Chain PSAD_BLOCK_OUTPUT (1 references) pkts bytes target prot opt in out source destination 0 0 DROP all -- * * 0.0.0.0/0 5.6.7.8 Chain PSAD_BLOCK_FORWARD (1 references) pkts bytes target prot opt in out source destination 0 0 DROP all -- * * 0.0.0.0/0 5.6.7.8 0 0 DROP all -- * * 5.6.7.8 0.0.0.0/0 You will even receive an email alerts that inform you of the scan detection, as well as an email informing you that iptables rules have been added to auto- block the nmap_scanner! Wrapping It All Up +--------------------------------------------------------------------------- Congratulations, you've successfully implemented psad to actively detect and respond to signature Nmap scans! Keep in mind this is one of the more basic setups for psad. You can go even further and adjust danger levels to suit degrees of paranoia, put psad into forensics mode, incorporate the software with DShield, and even manually use psad to manipulate iptables rules. A great resource for psad research is 'Linux Firewalls' by Michael Rash. Rash includes several chapters on psad covering not only theory but advanced implementation of psad from start to finish. If you wish to gain suggestions for an advanced, finely-tuned active defense setup with psad, be sure to check this book out! Have fun implementing an active defense against those who try to scan your system! Resources +--------------------------------------------------------------------------- / https://guardiandigital.com/ 'Linux Firewalls' by Michael Rash 'Knock, Knock, Knockin' on EnGarde's Door' features/features/knock-knock-knockin-on-engardes-door-with-fwknop
