Showing posts with label Attack. Show all posts
Showing posts with label Attack. Show all posts

ARP ATTACKS, ARP DOS AND INSPECTION


ARP ATTACKS AND INSPECTION

           Ø  Inside host of Firewall will exchanges ARP packets with unprotected hosts on the outside.
           Ø  A malicious host, however, can abuse ARP to leverage a man-in-the-middle attack.
           Ø  Example : Host 192.168.100.222 on the inside of the ASA is configured to use 192.168.100.1 as its default gateway.
           Ø  When the inside host needs to communicate with a destination that is outside of subnet 192.168.100.0/24
           Ø  It must send those packets to the gateway, which will relay the packets on toward their destination.
           Ø  The inside host does this by sending the packets to the gateway router’s MAC address.







           Ø  Now suppose that a malicious host exists on the unprotected outside part, and would like to insert itself in the middle of the inside host’s conversation so it intercept and examine the traffic.
           Ø  When the ARP request packet is forwarded outside then attacker catch it and reply his own MAC address with IP address.
           Ø  Now by this host came under Man-In-The-Middle attack.

ARP-Inspection

           Ø  To detect and prevent ARP spoofing, you can configure the ASA to support ARP inspection.
           Ø  ARP inspection uses static ARP entries as the basis for its inspection process.
           Ø  The ASA will examines each ARP reply packet it overhears and compare the source IP and MAC address, as well as the source interface, to known static entries in its own ARP table.
           Ø  If the ARP reply matches an existing entry, the ASA forwards it out the other interface else drop it.


Define the ARP configuration

           Ø  For ARP inspection, we define static ARP entries for all of the known hosts:

                      Ciscoasa(config)# arp  <interface>  <ip-address>  <mac-address>

            Ø  Static ARP entries never age out. To clear a static ARP entry, repeat the command by beginning with the “no” keyword.
            Ø  Enable ARP-inspection by command:

                      Ciscoasa(config)# arp-inspection  <interface> enable  [ flood | no-flood ]
 
              Ø  As ARP reply packets are received, the ASA will take one of the following actions:

·         If the MAC address and the IP address are both found in a single ARP table entry, the ARP reply must be valid and therefore is allowed to pass through the ASA.

·         If either the MAC address or the IP address is found in  ARP table, bit not both in a single entry, the ARP reply contains invalid or spoofed information. Therefore, it is dropped and not forwarded through the ASA.

·         If neither MAC address nor IP address is found in the ARP table, we can select the action as either flood (forward ot flood the ARP packet out the other firewall interface so it can reach its destination) or no-flood (drop the ARP packet without forwarding) . By default flood action is assumed.












Disabling MAC-Address Learning:

            Ø  In transparent Firewall mode, an ASA will learn MAC addresses as they are received on either of its interfaces.
            Ø  Suppose a malicious host is located on the outside of the ASA, as below figure.
            Ø  Before the malicious host begins it work, the inside host has sent a packet and the ASA has learned that its source MAC address (0000.2222.2222) is located on the inside interface.
            Ø  Now malicious host begin sending its own packets, but with spoofed MAC addresses in the source MAC address field.
            Ø  By sending a packet with source address 0000.2222.2222, the host can force the ASA to learn what is actually the inside host’s address on the outside interface.
            Ø  At this point, any packets arriving from the outside and destined for the inside host will simply be drooped because the host’s MAC address has been learned on the outside.









            Ø  So far so good, but the malicious host might not stop with just one spoofed MAC address. It might also send so many packets with spoofed addresses that ASA’s MAC table is overrun.
            Ø  That meant it’s a DOS attack on the ASA, rendering it unable to forward any packets correctly.
            Ø  To prevent by it , disable MAC address learning dynamically. And firewall must use statically configured MAC address entries.
            Ø  The command to disable learning:

                          Ciscoasa(config)# mac-learn <interface> disable

            Ø  Now configure a static MAC address entry by also specify the interface and MAC:

                          Ciscoasa(config)# mac-address-table static <interface mac-address>

By Er.AJAI SINGH on Monday, 4 February 2013 | , , | A comment?

Using uRPF at the access layer against DDoS attacks


How do you protect your devices from unauthorized Vitrual Terminal Connection?
Ø  Common way, simple inbound ACL applied to the VTY line or control-plane, like:
!
# ip access-list 10 permit 10.11.11.0 0.0.0.255
!
#line vty 0 4
   #access-class 10


TCP SYN Attack:

Ø  A TCP SYN flood attacks is accomplished by transmitting an TCP SYN packets to a host to exhaust its incoming TCP connection Queue.
Ø  The Size of this queue on TCP/IP stack implementation is small and easily exhausted.
Ø  Unicast RPF provides a source validation step to packet handling; it does this by verifying the source information of a packet to inform contained within the cisco express forwarding (CEF) forwarding information base (FIB).
Ø  The CEF FIB is a table that contains packet-switching information that mirrors that of the routing table; this is used by the device to increase the speed of pkts being forwarding through the device.
Ø  CEF must be configured on the device before uRPF is configured.
Ø  Unicast RPF operates in one of two modes:
·         Strict (normal):  When in strict mode, Unicast verifies that the source address in the FIB and that source address was received on the best return route interface as determined by CEF.
-          Can also be troublesome if routing is multihomed.
-          This is because the best return path might not be the same as the receiving interface ; because of this, strict uRPF is limited to single-homed connection.
-          uRPF will also work in situations where there are multiple equal-metric best paths available.
·         Loose:  Loose mode verifies only that the source address exists within the FIB and not the interface.
-          Loose mode allows additional flexivility to implement uRPF in locations where multihoming is common .
Ø  Another important things to understand about uRPF is that it only works on incoming interface.
Ø  If a single-homed connection existed between the network and an ISP, RPF would be configured to monitor traffic coming from the ISP only.


Unicast RPF Configuration:

Ø  The first thing that must be configured before starting Unicast RPF conf is to enable the use of CEF.
Ø  The ip cef {distributed} command enables the use of CEF

# ip cef { distributed }

Ø  The next part requires enabling uRPF on the incoming interface.
Ip verify unicast source reachable-via [rx | any ] {access-list}

Ø  Rx is used for Strict mode and any is used for loose mode.
Ø  An access-list can specified, its purpose is to determine whether the traffic will be dropped (default behavior-deny) or forwarded (permit).
Ø  It is important to understand that this access-list is not consider unless the pkt fails the unicast RPF check.

Verification:::::::
# show cef interface <interface>
To verify global uRPF pkt count,
# show ip traffic
To verify the number of interface uRPF pkt drops :
# show ip interface <interface>

By Er.AJAI SINGH on Sunday, 20 January 2013 | , | A comment?