Netsec WS 20131109 BD

Network Security
Workshop

Dhaka, Bangladesh
9 - 11 November 2013
Proudly Supported by:
Presenter
Nurul Islam (Roman)
Senior Training Specialist, APNIC
Nurul maintains the APNIC training lab and is involved in delivering
technical training for the APNIC community. He possesses specialized
skills in designing and running IPv4/IPv6 routing and switching
infrastructure for service provider and enterprise networks. Prior to his
current role he looked after the IP and AS number allocations for the
APNIC Members.

Areas of interests:
Internet Resource Management, IPv6, Routing and Switching, MPLS,
BGP, Security, Internet Routing Registry and RPKI, ISP Services and
Internetworking.

Contact:
Email: [email protected]

Presenter
Sheryl Hermoso (Shane)
Training Officer, APNIC
Sheryl has had various roles as a Network and Systems
Administrator prior to joining APNIC. She started her career as a
Technical Support Assistant while studying at the University of the
Philippines. She later finished her degree in Computer
Engineering and continued to work in the same university as a
Network Engineer, where she managed the DILNET network
backbone and wireless infrastructure.
Areas of interests:
DNS/DNSSEC, IPv6, Network Security, Internet Resource
Management

Contact:
Email: [email protected]

Overview
Network Security Fundamentals
Security on Different Layers and Attack Mitigation
Device and Infrastructure Security
Cryptography and PKI
Virtual Private Networks and IPsec
Resource Registration
RPKI and Resource Certification
Route Filtering
DNSSEC
Network Security
Fundamentals
Network Security Workshop
Overview
Why We Need Security
Definitions and Concepts
Access Control
Risk vs. Vulnerability
Threats and Attack Types

Why Security?
The Internet was initially designed for connectivity
Trust assumed
We do more with the Internet nowadays
Security protocols are added on top of the TCP/IP
Fundamental aspects of information must be protected
Confidential data
Employee information
Business models
Protect identity and resources
We cant keep ourselves isolated from the Internet
Most business communications are done online
We provide online services
We get services from third-party organizations online

Internet Evolution
Different ways to handle security as the Internet evolves
LAN connectivity
Application-specific
More online content
Application/data
hosted in the cloud
Attack Trends
Key findings:
Hacktivism and vandalism are the common DDoS attack motivation
High-bandwidth DDoS attacks are the new normal
First-ever IPv6 DDoS attacks are reported in 2011
Trust issues across geographic boundaries
Source: Arbor Networks Worldwide Infrastructure
Security Report Volume VIII
Attack Trends
Use of Distributed Reflection Denial of Service (DrDoS)
attacks
Shift away from SYN floods to UDP-based attacks
Chargen Protocol (UDP port 19) for DrDoS attacks
Infrastructure-directed attacks (L3 and L4)
For Q3 alone, the peak
Bandwidth average: 3.06 Gbps
Packets per second (PPS): 4.22 Mpps
Duration: 21.33 hours (38 hrs in Q2)
Theres a heightened level of global DDoS attack activity
Source: Prolexic Q3 2013 Global DDOS Attack Report
Attack Trends - Breach Sources
Infiltration
Aggregation
Exfiltration
Source: Trustwave 2012 Global Security Report
Goals of Information Security
Confidentiality Integrity Availability
S
E
C
U
R
I
T
Y

prevents
unauthorized
use or
disclosure of
information
safeguards the
accuracy and
completeness
of information
authorized
users have
reliable and
timely access
to information
Access Control
The ability to permit or deny the use of an object by a
subject.
It provides 3 essential services:
Authentication (who can login)
Authorization (what authorized users can do)
Accountability (identifies what a user did)

Authentication
A means to verify or prove a users identity
The term user may refer to:
Person
Application or process
Machine or device
Identification comes before authentication
Provide username to establish users identity
To prove identity, a user must present either of the following:
What you know (passwords, passphrase, PIN)
What you have (token, smart cards, passcodes, RFID)
Who you are (biometrics such as fingerprints and iris scan, signature or
voice)
Examples of Tokens
eToken
RFID cards
Smart Cards
Fingerprint scanner
Trusted Network
Standard defensive-oriented technologies
Firewall first line of defense
Intrusion Detection second line of defense
Build TRUST on top of the TCP/IP infrastructure
Strong authentication
Two-factor authentication
something you have + something you know
Public Key Infrastructure (PKI)
Strong Authentication
An absolute requirement
Two-factor authentication
Passwords (something you know)
Tokens (something you have)
Examples:
Passwords
Tokens
Tickets
Restricted access
PINs
Biometrics
Certificates

Two-factor Authentication
Requires a user to provide at least two authentication
factors to prove his identity
something you know
Username/userID and password
something you have
Token using a one-time password (OTP)
The OTP is generated using a small electronic device in
physical possession of the user
Different OTP generated each time and expires after some time
An alternative way is through applications installed on your mobile
device
Multi-factor authentication is also common
Authorization
Defines the users rights and permissions on a system
Typically done after user has been authenticated
Grants a user access to a particular resource and what
actions he is permitted to perform on that resource
Access criteria based on the level of trust:
Roles
Groups
Location
Time
Transaction type
Authentication vs. Authorization
Client
Service
Authentication
Mechanism
Authorization
Mechanism
Authentication simply identifies a party, authorization defines whether they can
perform certain action RFC 3552
Authorization Concepts
Authorization creep
When users may possess unnecessarily high access privileges within
an organization
Default to Zero
Start with zero access and build on top of that
Need to Know Principle
Least privilege; give access only to information that the user
absolutely need
Access Control Lists
List of users allowed to perform particular access to an object (read,
write, execute, modify)

Single Sign On
Property of access control where a user logs in only once
and gains access to all authorized resources within a
system.
Benefits:
Ease of use
Reduces logon cycle (time spent re-entering passwords for the same
identity)
Common SSO technologies:
Kerberos, RADIUS
Smart card based
OTP Token
Disadvantage: Single point of attack
Types of Access Control
Centralized Access Control
Radius
TACACS+
Diameter
Decentralized Access Control
Control of access by people who are closer to the resources
No method for consistent control
Accountability
The security goal that generates the requirement for actions
of an entity to be traced uniquely to that entity
Senders cannot deny sending information
Receivers cannot deny receiving it
Users cannot deny performing a certain action
Supports nonrepudiation, deterrence, fault isolation,
intrusion detection and prevention and after-action recovery
and legal action

Source: NIST Risk Management Guide for
Information Technology Systems
Integrity
Security goal that generates the requirement for protection
against either intentional or accidental attempts to violate
data integrity
Data integrity
The property that data has when it has not been altered in an
unauthorized manner
System integrity
The quality that a system has when it performs its intended function
in an unimpaired manner, free from unauthorized manipulation
Source: NIST Risk Management Guide for
Information Technology Systems
Risk, Threat and Vulnerability
Vulnerability - weakness in a system
Risk - likelihood that a particular threat using a particular
attack will exploit a particular vulnerability
Exploit - taking advantage of a vulnerability
Non-repudiationassurance that both parties are involved
in the transaction

Vulnerability
A weakness in security procedures, network design, or
implementation that can be exploited to violate a corporate
security policy
Software bugs
Configuration mistakes
Network design flaw
Lack of encryption
Exploit
Taking advantage of a vulnerability
Threat
Any circumstance or event with the potential to cause harm
to a networked system
a motivated, capable adversary
These are some example of threats:
Denial of service
Attacks make computer resources (e.g., bandwidth, disk space, or CPU time)
unavailable to its intended users
Unauthorised access
Access without permission issues by a rightful owner of devices or networks
Impersonation
Worms
Viruses
Attack Sources
Active vs. passive
Active involves writing data to the network. It is common to disguise
ones address and conceal the identity of the traffic sender.
Passive involves only reading data on the network. Its purpose is breach
of confidentiality. This is possible if:
Attacker has gained control of a host in the communication path between two victim
machines
Attacker has compromised the routing infrastructure to arrange the traffic pass through a
compromised machine
Active Attacks Passive Attacks
Denial of Service attacks
Spoofing
Man in the Middle
ARP poisoning
Smurf attacks
Buffer overflow
SQL Injection
Reconnaissance
Eavesdropping
Port scanning

Source: RFC 4778
Attack Sources
On-path vs. Off-path
On-path routers (transmitting datagrams) can read, modify, or remove
any datagram transmitted along the path
Off-path hosts can transmit datagrams that appear to come from any
hosts but cannot necessarily receive datagrams intended for other
hosts
If attackers want to receive data, they have to put themselves on-path
How easy is it to subvert network topology?
It is not easy thing to do but, it is not impossible
Insider vs. outsider
What is definition of perimeter/border?
Deliberate attack vs. unintentional event
Configuration errors and software bugs are as harmful as a
deliberate malicious network attack
Source: RFC 4778
General Threats
Masquerade
An entity claims to be another entity
Eavesdropping
An entity reads information it is not intended to read
Authorization violation
An entity uses a service or resource it is not intended to use
Loss or modification of information
Data is being altered or destroyed
Denial of communication acts (repudiation)
An entity falsely denies its participation in a communication act
Forgery of information
An entity creates new information in the name of another entity
Sabotage
Any action that aims to reduce the availability and/or correct functioning of services or
systems
Reconnaissance Attack
Unauthorised users to gather information about the network
or system before launching other more serious types of
attacks
Also called eavesdropping
Information gained from this attack is used in subsequent
attacks (DoS or DDoS type)
Examples of relevant information:
Names, email address
Common practice to use a persons first initial and last name for accounts
Practically anything

Man-in-the-Middle Attack
Active eavesdropping
Attacker makes independent connections with victims and
relays messages between them, making them believe that
they are talking directly to each other over a private
connection, when in fact the entire conversation is
controlled by the attacker
Usually a result of lack of end-to-end authentication
Masquerading - an entity claims to be another entity

Session Hijacking
Exploitation of a valid computer session, to gain
unauthorized access to information or services in a
computer system.
Theft of a magic cookie used to authenticate a user to a
remote server (for web developers)
Four methods:
Session fixation attacker sets a users session id to one known to
him, for example by sending the user an email with a link that
contains a particular session id.
Session sidejacking attacker uses packet sniffing to read network
traffic between two parties to steal the session cookie.
Denial of Service (DoS) Attack
Attempt to make a machine or network resource unavailable to
its intended users.
Purpose is to temporarily or indefinitely interrupt or suspend
services of a host connected to the Internet
Methods to carry out this attack may vary
Saturating the target with external communications requests (such that it
cant respond to legitimate traffic) SERVER OVERLOAD
May include malware to max out target resources (such as CPU), trigger
errors, or crash the operating system
DDoS attacks are more dynamic and comes from a broader
range of attackers
Examples: SYN flooding, Smurf attacks, Starvation
Can be used as a redirection and reconnaissance technique

Questions
Security in Layers &
Attack Mitigation
Overview
Attacks in Different Layers
Security Technologies
Link-Layer Security
Network Layer Security
Transport Layer Security
Application Layer Security

Attacks on Different Layers
Application
Presentation
Session
Transport
Network
Data Link
Physical
Application

Transport
Internet
Network Access
(Link Layer)
Layer 2: PPTP, Token Ring
Layer 3: IPv4, IPv6, ICMP, IPsec
Layer 4: TCP, UDP
Layer 5: SMB, NFS, Socks
Layer 7: DNS, DHCP, HTTP, FTP, IMAP,
LDAP, NTP, Radius, SSH, SMTP, SNMP,
Telnet, TFTP
Ping/ICMP Flood
TCP attacks, Routing
attack, SYN flooding,
Sniffing
DNS Poisoning, Phishing,
SQL injection, Spam/Scam
ARP spoofing, MAC
flooding
OSI Reference Model TCP/IP Model
Layer 2 Attacks
ARP Spoofing
MAC attacks
DHCP attacks
VLAN hopping

ARP Spoofing
1
Since the clients ARP Cache has
already been poisoned, it will
communicate directly to the fake
Destination.

I want to connect to
10.0.0.3. I dont know
the MAC address
10.0.0.1
AA-AA-AA-AA-AA-AA
10.0.0.2
BB-BB-BB-BB-BB-BB
10.0.0.3
CC-CC-CC-CC-CC-CC
10.0.0.4
DD-DD-DD-DD-DD-
DD
ARP Request
ARP Reply
Wait, I am 10.0.0.3!
I am 10.0.0.3. This is
my MAC address
MAC Flooding
Exploits the limitation of all switches fixed CAM table size
CAM = Content Addressable memory = stores info on the
mapping of individual MAC addresses to physical ports on
the switch.
Port 1 Port
2
Port
3
Port
4
00:01:23:45:67:A1 x
00:01:23:45:67:B2 x
00:01:23:45:67:C3 x
00:01:23:45:67:D4 x
DHCP Attacks
DHCP Starvation Attack
Broadcasting vast number of DHCP requests with spoofed MAC
address simultaneously.
DoS attack using DHCP leases
Rogue DHCP Server Attacks

Attacker sends many different DHCP requests
with many spoofed addresses.
Server runs out of IP addresses
to allocate to valid users
Layer 3 Attacks
ICMP Ping Flood
ICMP Smurf
Ping of death

Ping Flood
Internet
Broadcas
t Enabled
Network
Victim
1
Other forms of ICMP attack:
-Ping of death
-ICMP ping flood
Attacker
Mitigating Sniffing Attacks
Avoid using insecure protocols like basic HTTP
authentication and telnet.
If you have to use an insecure protocol, try tunneling it
through something to encrypt the sensitive data.
Run ARPwatch.
Try running tools like sniffdet and Sentinel to detect
network cards in promiscuous mode that may be running
sniffing software.
Routing Attacks
Attempt to poison the routing information
Distance Vector Routing
Announce 0 distance to all other nodes
Blackhole traffic
Eavesdrop
Link State Routing
Can drop links randomly
Can claim direct link to any other routers
A bit harder to attack than DV
BGP attacks
ASes can announce arbitrary prefix
ASes can alter path
TCP Attacks
SYN Flood occurs when an attacker sends SYN requests
in succession to a target.
Causes a host to retain enough state for bogus half-
connections such that there are no resources left to
establish new legitimate connections.
TCP Attacks
Exploits the TCP 3-way handshake
Attacker sends a series of SYN packets without replying
with the ACK packet
Finite queue size for incomplete connections
Server
CONNECTION ESTABLISHED
SYN
SYN+ACK
ACK
TCP Attacks
Exploits the TCP 3-way handshake
Attacker sends a series of SYN packets without replying
with the ACK packet
Finite queue size for incomplete connections
Server
(Victim)
Attacker
OPEN CONNECTIONS
SYN
SYN+ACK
ACK?
Application Layer Attacks
Applications dont authenticate properly
Authentication information in clear
FTP, Telnet, POP
DNS insecurity
DNS poisoning
DNS zone transfer
1
Scripting vulnerabilities
Cookie poisoning
Buffer overflow
Hidden field manipulation
Parameter tampering
Cross-site scripting
SQL injection

1
Source: Arbor Networks Worldwide Infrastructure
Security Report Volume VII
Application Layer DDoS: Slowloris
Incomplete HTTP requests
Properties
Low bandwidth
Keep sockets alive
Only affects certain web servers
Doesnt work through load balancers
Managed to work around accf_http
Web Application Security Risks
Injection
Cross-Site Scripting
Broken authentication and Session Management
Insecure Direct Object References
Cross-site Request Forgery (CSRF)
Insecure Cryptographic Storage
Failure to Restrict URL Access
Insufficient Transport Layer Protection
Unvalidated Redirects and Forwards
Source: OWASP Top 10 Application Security Risks, 2010
DNS Changer
Criminals have learned that if they can control a users
DNS servers, they can control what sites the user connects
to the Internet.
How: infect computers with a malicious software (malware)
This malware changes the users DNS settings with that of
the attackers DNS servers
Points the DNS configuration to DNS resolvers in specific
address blocks and use it for their criminal enterprise
For more: see the NANOG presentation by Merike

Rogue DNS Servers
85.225.112.0 through 85.255.127.255
67.210.0.0 through 67.210.15.255
93.188.160.0 through 93.188.167.255
77.67.83.0 through 77.67.83.255
213.109.64.0 through 213.109.79.255
64.28.176.0 through 64.28.191.255

If your computer is configured with one of these DNS
servers, it is most likely infected with DNSChanger malware
DNS Cache Poisoning
Caching incorrect resource record that did not originate
from authoritative DNS sources.
Result: connection (web, email, network) is redirected to
another target (controlled by the attacker)

DNS Cache Poisoning
(pretending to be
the authoritative
zone)
ns.example.com
Webserver
(192.168.1.1)
DNS Caching
Server
Client
I want to access
www.example.com
1
QID=6457
1
2
QID=64569
QID=64570
QID=64571
www.example.com 192.168.1.1
match!
3
3
Root/GTLD
QID=64571
DNS Amplification
A type of reflection attack combined with amplification
Source of attack is reflected off another machine
Traffic received is bigger (amplified) than the traffic sent by the
attacker
UDP packets source address is spoofed
DNS Amplification Attack
Queries for
www.example.com
Attacker
ns.example.com
Victim Server
DNS Recursive server
Compromised
Machines
(spoofed IP)
Root/GTLD
Common Types of Attack
Ping sweeps and port scans - reconnaissance
Sniffing capture packet as they travel through the network
Man-in-the-middle attack intercepts messages that are
intended for a valid device
Spoofing - sets up a fake device and trick others to send
messages to it
Hijacking take control of a session
Denial of Service (DoS) and Distributed DoS (DDoS)
Wireless Attacks
WEP first security mechanism for 802.11 wireless
networks
Weaknesses in this protocol were discovered by Fluhrer,
Mantin and Shamir, whose attacks became known as FMS
attacks
Tools were developed to automate WEP cracking
Chopping attack were released to crack WEP more
effectively and faster
Cloud-based WPA cracker
https://www.wpacracker.com/
Man in the Middle Attacks (Wireless)
Creates a fake access point and have clients authenticate
to it instead of a legitimate one.
Capture traffic to see usernames, passwords, etc that are
sent in clear text.

Botnet
Collection of compromised computers (or bot)
Computers are targeted by malware (malicious software)
Once controlled, an attacker can use the compromised
computer via standards-based network protocol such as
IRC and HTTP
How to become a bot:
Drive-by downloads (malware)
Go to malicious websites (exploits web browser vulnerabilities)
Run malicious programs (Trojan) from websites or as email
attachment
Password Cracking
Dictionary attacks
Guessing passwords using a file of 1M possible password values
Ordinary words and peoples names
Offline dictionary attack when the entire password file has been
attacked
Use random characters as password with varying upper and lower
case, numbers, and symbols
Brute-force attacks
Checking all possible values until it has been found
The resource needed to perform this attack grows exponentially while
increasing the key size
Social engineering

Pharming and Phishing
Phishing victims are redirected to a fake website that
looks genuine. When the victim supplies his account and
password, this can be used by the attacker to the target site
Typically uses fraud emails with clickable links to fake websites
Pharming redirect a websites traffic to another fake site
by changing the victims DNS settings or hosts file

Security on Different Layers
Application
Presentation
Session
Transport
Network
Data Link
Physical

Layer 2: ARP, Token Ring
Layer 3: IPv4, IPv6, ICMP, IPsec
Layer 4: TCP, UDP
Layer 5: SMB, NFS, Socks
Layer 7: DNS, DHCP, HTTP, FTP,
IMAP, LDAP, NTP, Radius, SSH,
SMTP, SNMP, Telnet, TFTP
Ping/ICMP Flood
TCP attacks, Routing attack,
SYN flooding, Sniffing
DNS Poisoning, Phishing, SQL
injection, Spam/Scam
ARP spoofing, MAC
flooding
IEEE 802.1X, PPP &
PPTP
IPsec
TLS, SSL, SSH

HTTPS, DNSSEC, PGP, SMIME

Link-Layer Security
Layer 2 Forwarding (L2F)
Point-to-Point Tunneling Protocol (PPTP)
Layer 2 Tunneling Protocol (L2TP)

Layer 2 Forwarding Protocol
Created by Cisco Systems and replaced by L2TP
Permits the tunneling of the link layer High-level Data Link
Control (HDLC), async HDLC, or Serial Line Internet
Protocol (SLIP) frames of higher-level protocols
IP/UDP L2F PPP (Data)
Carrier
Protocol
Encapsulator
Protocol
Passenger
Protocol
Point to Point Tunneling Protocol
Initiated by Microsoft but later became an informational
standard in the IETF (RFC 2637)
Client/server architecture that allows PPP to be tunneled
through an IP network and decouples functions that exist in
current NAS.
Connection-oriented
Layer 2 Tunneling Protocol
Combination of L2F and PPTP
Published as RFC 2661 and known as L2TPv2
L2TPv3 provides additional security features and the ability
to carry data links other than PPP
The two end-points are L2TP Access Concentrator (LAC) or
L2TP Network Server (LNS)

PPPoE
PPP over Ethernet
Defined in RFC 2516
A means to encapsulate PPP packets over the Ethernet link
layer
Mostly used in ADSL environments to provide access
control, billing, and type of service on a per-user rather than
a per-site basis
Transport Layer Security
Secure Socket Layer (SSL)
Secure Shell Protocol
SOCKS Protocol
SSL/TLS
TLS and SSL encrypts the segments of network
connections above the Transport Layer.
Versions:
SSLv1 designed by Netscape
SSLv2 publicly released in 1994; has a number of security flaws;
uses RC4 for encryption and MD5 for authentication
SSLv3 added support for DSS for authentication and DH for key
agreement
TLS based on SSLv3; uses DSS for authentication, DH for key
agreement, and 3DES for encryption
TLS is the IETF standard which succeeded SSL.
SSL Handshake
Client initiates SSL connection
Server sends digital certificate
and selected cipher suite
Client sends encrypted shared secret
Message encryption and integrity algorithms
are negotiated
Session keys are generated
k
bs
Advantages of SSL
The connection is private
Encryption is used after initial handshake to define a secret key
Encryption uses symmetric cryptography (DES or RC4)
Peers identity can be authenticated using asymmetric
cryptography (RSA or DSS)
The connection is reliable
Message transport includes message integrity check using a keyed
MAC. Secure hash functions (SHA or MD5) are used for MAC
computation.
Applications Using SSL/TLS
Protocol Defined Port Number SSL/TLS Port Number
HTTP 80 443
NNTP 119 563
LDAP 389 636
FTP-data 20 989
FTP-control 21 990
Telnet 23 992
IMAP 143 993
POP3 110 995
SMTP 25 465
Secure Shell Protocol (SSH)
Protocol for secure remote login
Provides support for secure remote login, secure file
transfer, and secure forwarding of TCP/IP and X Window
System traffic
Consists of 3 major components:
Transport layer protocol (server authentication, confidentiality,
integrity)
User authentication protocol (authenticates client to the server)
Connection protocol (multiplexes the encrypted tunnel into several
logical channels)
Application Layer Security
HTTPS
PGP (Pretty Good Privacy)
SMIME (Secure Multipurpose Internet Mail Extensions)
TSIG and DNSSEC
Wireless Encryption - WEP, WPA, WPA2
HTTPS
Hypertext Transfer Protocol Secure
Widely-used, message-oriented communications protocol
Connectionless oriented protocol
Technically not a protocol in itself, but simply layering HTTP
on top of the SSL/TLS protocol
Encapsulates data after security properties of the session
Not to be confused with S-HTTP
Note: A website must use HTTPS everywhere, otherwise it is still
vulnerable to some attacks
Pretty Good Privacy (PGP)
Stands for Pretty Good Privacy, developed by Phil
Zimmerman in 1995
PGP is a hybrid cryptosystem
combines some of the best features of both conventional and public
key cryptography
Assumptions:
All users are using public key cryptography and have generated
private/public key pairs (using RSA or El Gamal)
All users also use symmetric key system (DES or Rijndael)
Offers authentication, confidentiality, compression, e-mail
compatibility and segmentation
Questions
Device & Infrastructure
Security
Overview
Server Hardening
Logging and Monitoring
Backup and Recovery
L2 Device (Switch) Security
L3 Device (Router) Security
Bogons
Server Hardening
Use netstat to check which ports you are currently listening
on your Linux machine.
Close unnecessary ports
Remove unused applications. Minimal software means less
possible vulnerabilities
Perform regular software patches and update.
Disable unwanted services and remove from startup items
Use TCP wrappers and properly configure hosts.allow and
hosts.deny files.
Rule of thumb: deny all, allow as necessary

Server Hardening - Accounts
Disable default accounts and groups that are not needed
Use strong authentication
Good password policy
Minimum number of characters
Combination of alphabets, numbers, special characters, upper and
lower case
Implement password aging
Force users to change password on first login
Prevent use of previous passwords
Lock account after a number of failed logins

Backup and Recovery
A backup system is a fundamental element of any disaster
recovery plan
Provides for disaster recovery of key network services and
any file
Physical vs Logical backup
Physical backup create copy of the files into some other location,
such as disks or tapes
Logical backup remote, cloud services
Backup and Recovery
Type of Backup
Full backup entire dataset regardless whether it has been altered or
not; take longer
Incremental backups only backs up data that has been changed
since last backup
Differential backups backups data that has changed since the last
full backup
Copy backups full backups without a reset of archive bits
Backup intervals
Daily incremental
Backup media
Offline backups: Magnetic Tapes, Optical drives, Hard disk
Offsite backups: Network, NAS, Data center
Backup and Recovery
Always test that your backups are restorable
Restoration should be within a reasonable period of time
time to recover
Priority of some files over others

Logging
All security-related events must be logged
Audits must be performed on a regular basis
In a Linux system:
monitor log messages using logwatch
Perform system accounting using auditd

Network Devices
Attack areas:
Default passwords
Broadcasting packets replicated on
all ports
rogue devices participate in STP
and mislead it
packet flooding using spoofed
MAC addresses
Gaining management access
Network Devices
Change the default settings
Allow management sessions only from approved sources
Use AAA server to authenticate administrators, authorize
their actions and perform accounting of all actions
Encrypt sessions
Limit device access
Console sessions should also be authenticated
Admin actions should be authorized through a AAA server
Disable password recovery

Increasing Port Security
CAM entry aging
Static or permanent CAM entries
Limit MAC addresses per port
Disable unused port
Port authentication with IEEE 802.1X
Mitigation of ARP Spoofing
Static ARP entries
Secure ARP inspection
Secure ARP Discovery (SAD)
Separate VLANs with unique IP subnets
Private VLAN
IEEE 802.1X port authentication

Port Authentication using 802.1X
802.1X is an IEEE standard defining layer 2 protocol used
for authentication purposes
3 authentication protocols
EAP-MD5 (challenge response auth protocol)
EAP-OTP (proprietary one-time password)
EAP-TLS (using digital certificates)
Spanning Tree Protocol
Used to prevent loops in a switched LAN by disabling
redundant links (IEEE 802.1D)
STP takes 30-50 seconds to converge
Misplacement of the root bridge can cause suboptimal
paths
Loss of the root bridge has the most effect on performance
STP does not include any security by default
Trunking
Allows multiple VLANs to be carried over one physical link.
This may expose the whole network to a single port if not
configured properly.
An intruder can also interfere with VTP messages and
cause VLANs to disappear
Must disable trunking on a port
VLAN Trunking Protocol (VTP)
used to make VLAN management easier in large switched
LAN environments
VTP should use authentication to prevent unauthorised
devices from participating in VTP
VTP can use MD5 to authenticate VTP messages
The same password has to be used on all devices in a VTP
domain
Replay attacks are not possible
VLAN/Trunking Best Practices
Always use a dedicated VLAN ID for all trunk ports
Disable unused ports and put them in an unused vlan
Do not use VLAN 1 for anything
Disable auto-trunking on user facing ports (DTP off)
Explicitly configure trunking on infrastructure ports
Use all tagged mode for the native VLAN on trunks
Use PC Voice VLAN access on phones that support it
Use 802.1q tag on all the trunk ports

VLAN 1
A special VLAN used as a default vlan assigned to L2
device ports.
Generic rule: network administrators should pruse any
VLAN, and in particular VLAN 1, from all theports where
that VLAN is not strictly needed
Network Devices: Routers
Use strong authentication
Disable unused services
Modify insecure default settings
Authenticate IGP messages
Check software versions for security-related bugs
Network Hardening (Layer 3)
Interior gateway protocol (IGP) - exchange routing
information between routers inside the network
OSPF, EIGRP, RIPv2, IS-IS
Exterior Gateway Protocol (EGP) - exchange routes with
Internet Service Providers (ISPs)
BGP
Threats to Routing Protocols
Deliberate exposure attacker takes control of a router and
intentionally releases routing information to other entities
Sniffing attackers monitor and/or record the routing exchanges
between authorized routers to sniff for routing information
Traffic analysis attackers gain routing information by analyzing
the characteristics of the data traffic on a subverted link
Spoofing illegitimate device assumes the identity of a legitimate
one.
Falsification attacker sends false routing information
Interference attacker inhibits the exchanges by legitimate
routers
Overload attackers place excess burden on legitimate routers
RFC 4593
Securing Routing Protocols
an authentication mechanism should be used to prevent
accidental or deliberate adjacencies from being established
password system should be secret and changed regularly
authenticate routing updates
security only verifies the source of the information
no encryption of the routing update contents
any packet interception will allow read-access

SIDR Working Group
Secure Inter-Domain Routing (SIDR)
Its purpose is to reduce vulnerabilities to the inter-domain
routing system
Addresses two vulnerabilities:
Is an Autonomous System authorized to originate an IP prefix?
Is the AS-Path represented in the route the same as the path through
which the NLRI traveled?
Projects:
PKI, RPKI, BGPsec
Source: SIDR WG
https://datatracker.ietf.org/wg/sidr/charter/
What is RPKI?
Resource Public Key Infrastructure (RPKI)
A robust security framework for verifying the association between
resource holder and their Internet resources
Created to address the issues in RFC 4593
Uses X.509 v3 certificates
With RFC3779 extensions
Helps to secure Internet routing by validating routes
Proof that prefix announcements are coming from the legitimate holder
of the resource
A system to manage the creation and storage of digital
certificates and the associated Route Origin Authorization
documents

BGP Security (BGPsec)
Extension to BGP that provides improved security for BGP
routing
Currently an IETF Internet draft
Implemented via a new optional non-transitive BGP path
attribute that contains a digital signature
BGP Prefix Origin Validation (using RPKI)
BGP Path Validation
BGP Security
Real-time Blackhole Routing (RTBH)
Packets are forwarded to a routers bit bucket either a null interface or
a discard interface)
desired packets are dropped with minimal or no performance impact
Employing uRPF in conjunction with RTBH can provide source-based
solution vs destination-based
BGP Diversion
Uses BGP to divert traffic to sinkholes or any packet scrubbing centers
for further analysis
Divert via resetting BGP next hop to ip address of analysis system(s) or
matching community tags that result in different BGP next hops being
assigned for a given prefix
BGP Route Tagging
Tag routes using BGP communities to apply filtering, rate limiting, QoS,
firewall, or any other policy on packets

Sinkholes
Cisco IOS Features
Control Plane Policing
allows users to configure a QoS filter that manages the traffic flow of
control plane packets to protect the control plane of Cisco IOS
routers and switches against reconnaissance and denial-of-service
(DoS) attacks
Control plane can help maintain packet forwarding despite an attack
or heavy traffic load
Unicast RPF
limits malicious traffic by enabling a router to verify the reachability of
the source address in packets being forwarded.

Bogons
Bogons Martians and netblocks that have not been
allocated to an RIR by IANA.
Fullbogons a larger set which also includes IP space that
has been allocated to an RIR, but not assigned by that RIR
to an actual ISP or other end-users.

Reference: http://www.team-cymru.org/Services/Bogons/bgp.html
Bogons (BGP)
Cisco Router Traditional bogons
router bgp <your asn>
neighbor x.x.x.x remote-as 65333
neighbor x.x.x.x ebgp-multihop 255
neighbor x.x.x.x description <your description>
neighbor x.x.x.x prefix-list cymru-out out
neighbor x.x.x.x route-map CYMRUBOGONS in
neighbor x.x.x.x password <your password>
neighbor x.x.x.x maximum-prefix 100 threshold
90

Reference: http://www.team-cymru.org/Services/Bogons/bgp.html
Bogons (BGP)
! Remember to configure your Cisco router to handle the new style
! community syntax.
ip bgp-community new-format
!
! Set a bogon next-hop on all routers that receive the bogons.
ip route 192.0.2.1 255.255.255.255 null0
!
! Configure a community list to accept the bogon prefixes into the
! route-map.
ip community-list 10 permit 65333:888
!
! Configure the route-map. Remember to apply it to the proper
! peering sessions.
route-map CYMRUBOGONS permit 10
description Filter bogons learned from cymru.com bogon route-servers
match community 10
set ip next-hop 192.0.2.1
!
ip prefix-list cymru-out seq 5 deny 0.0.0.0/0 le 32
Bogons (BGP Peer-Group)
router bgp <your asn>
neighbor cymru-bogon peer-group
neighbor cymru-bogon ebgp-multihop 255
neighbor cymru-bogon description <general description>
neighbor cymru-bogon prefix-list cymru-out out
neighbor cymru-bogon route-map CYMRUBOGONS in
neighbor cymru-bogon maximum-prefix 100 threshold 90
! You'll need to increase the maximum to at least 50000 with
an
! appropriate thresholds if you're receiving one or both
fullbogons
! feeds.
!
neighbor x.x.x.x remote-as 65333
neighbor x.x.x.x peer-group cymru-bogon
neighbor x.x.x.x description <specific description>
neighbor x.x.x.x password <your password>
Questions
Cryptography

Overview
What is Cryptography?
Symmetric Key Cryptography
Asymmetric Key Cryptography
Block and Stream Cipher
Digital Signature and Message Digest
Cryptography
Cryptography is everywhere
German Lorenz cipher machine
Cryptography
Cryptography deals with creating documents that can be
shared secretly over public communication channels
Other terms closely associated
Cryptanalysis = code breaking
Cryptology
Kryptos (hidden or secret) and Logos (description) = secret speech /
communication
combination of cryptography and cryptanalysis
Cryptography is a function of plaintext and a cryptographic
key
C = F(P, k)
Notation:
Plaintext (P)
Ciphertext (C)
Cryptographic Key (k)
Typical Scenario
Alice wants to send a secret message to Bob
What are the possible problems?
Data can be intercepted
What are the ways to intercept this message?
How to conceal the message?
Encryption
Crypto Core
Secure key establishment

Secure communication

Alice has key (k)
Bob has key (k)
Alice has key (k)
Bob has key (k)
m
m
m
Confidentiality and integrity
Source: Dan Boneh, Stanford
It can do much more
Digital Signatures
Anonymous communication
Anonymous digital cash
Spending a digital coin without anyone knowing my identity
Buy online anonymously?
Elections and private auctions
Finding the winner without actually knowing individual votes (privacy)

Other uses are also theoretically
possible (Crypto magic)
Privately outsourcing computation

Zero knowledge (proof of knowledge)
Alice with
search query
Google
E(query)
E(results)
What did
she search
for?
I know the factorial of N Bob
I know the factorial of N
Proof
History: Ciphers
Substitution cipher
involves replacing an alphabet with another character of the same
alphabet set
Can be mono-alphabetic (single set for substitution) or poly-
alphabetic system (multiple alphabetic sets)
Example:
Caesar cipher, a mono-alphabetic system in which each character is
replaced by the third character in succession
Vigenere cipher, a poly-alphabetic cipher that uses a 26x26 table of
characters
How to Break a Substitution Cipher
UKBYBIPOUZBCUFEEBORUKBYBHOBBRFESPVKBWFOFERVNBCVBZPRUBOFERVNBCVBPCYYFVUFO
FEIKNWFRFIKJNUPWRFIPOUNVNIPUBRNCUKBEFWWFDNCHXCYBOHOPYXPUBNCUBOYNRVNIWN
CPOJIOFHOPZRVFZIXUBORJRUBZRBCHNCBBONCHRJZSFWNVRJRUBZRPCYZPUKBZPUNVPWPCYVF
ZIXUPUNFCPWRVNBCVBRPYYNUNFCPWWJUKBYBIPOUZBCUIPOUNVNIPUBRNCHOPYXPUBNCUB
OYNRVNIWNCPOJIOFHOPZRNCRVNBCUNENVVFZIXUNCHPCYVFZIXUPUNFCPWZPUKBZPUNVR

(1) Use frequency of the English letters
e = 12.7%
t = 9.1 %
a = 8.1%
(2) Use frequency of pairs of letters
he, in, an, th
In the example,
B appeared 36 times, U 33 times, and P 32 times
NC appeared 11 times, PU 10 times
UKB appeared 6 times
Transposition Cipher
No letters are replaced, they are just rearranged.
Rail Fence Cipher another kind of transposition cipher in
which the words are spelled out as if they were a rail fence.
History: Rotor Machines (1870-1943)
Hebern machine single rotor

Enigma - 3-5 rotors
Source: Wikipedia (image)
Modern Crypto Algorithms
specifies the mathematical transformation that is performed
on data to encrypt/decrypt
Crypto algorithm is NOT proprietary
Analyzed by public community to show that there are no
serious weaknesses
Explicitly designed for encryption

Encryption
process of transforming plaintext to ciphertext using a
cryptographic key
Used all around us
In Application Layer used in secure email, database sessions, and
messaging
In session layer using Secure Socket Layer (SSL) or Transport Layer
Security (TLS)
In the Network Layer using protocols such as IPsec
Benefits of good encryption algorithm:
Resistant to cryptographic attack
They support variable and long key lengths and scalability
They create an avalanche effect
No export or import restrictions
Encryption and Decryption
Plaintext
ENCRYPTION
ALGORITHM
DECRYPTION
ALGORITHM
Ciphertext
Plaintext
Encryption Key Decryption Key
Symmetric Key Algorithm
Uses a single key to both encrypt and decrypt information
Also known as a secret-key algorithm
The key must be kept a secret to maintain security
This key is also known as a private key
Follows the more traditional form of cryptography with key
lengths ranging from 40 to 256 bits.
Examples:
DES, 3DES, AES, RC4, RC6, Blowfish

Symmetric Encryption
Plaintext
ENCRYPTION
ALGORITHM
DECRYPTION
ALGORITHM
Ciphertext
Plaintext
Same shared secret key
Shared Key Shared Key
Symmetric Key
Cryptography
Symmetric Key Algorithm
Symmetric Algorithm Key Size
DES 56-bit keys
Triple DES (3DES) 112-bit and 168-bit keys
AES 128, 192, and 256-bit keys
IDEA 128-bit keys
RC2 40 and 64-bit keys
RC4 1 to 256-bit keys
RC5 0 to 2040-bit keys
RC6 128, 192, and 256-bit keys
Blowfish 32 to 448-bit keys
Note:
Longer keys are more difficult to crack, but more computationally expensive.
Data Encryption Standard (DES)
Developed by IBM for the US government in 1973-1974,
and approved in Nov 1976.
Based on Horst Feistels Lucifer cipher
block cipher using shared key encryption, 56-bit key length
Block size: 64 bits
DES: Illustration
Plaintext
ENCRYPTION
ALGORITHM
DECRYPTION
ALGORITHM
Ciphertext
Plaintext
56-bit keys +
8 bits parity
64-bit blocks of input text
Triple DES
3DES (Triple DES) a block cipher that applies DES three
times to each data block
Uses a key bundle comprising of three DES keys (K1, K2,
K3), each with 56 bits excluding parity.
DES encrypts with K1, decrypts with K2, then encrypts with
K3

Disadvantage: very slow
C
i
= E
K3
(D
K2
(E
K1
(P
i
)))
3DES: Illustration

Note:
If Key1 = Key2 = Key3, this is similar to DES
Usually, Key1 = Key3
Plaintext
ENCRYPT
Ciphertext
Key 1
DECRYPT ENCRYPT
Key 2 Key 3
Advanced Encryption Standard (AES)
Published in November 2001
Symmetric block cipher
Has a fixed block size of 128 bits
Has a key size of 128, 192, or 256 bits
Based on Rijndael cipher which was developed by Joan
Daemen and Vincent Rijmen
Better suited for high-throughput, low latency environments
Rivest Cipher
RC Algorithm Description
RC2 Variable key-sized cipher used as a drop in replacement
for DES
RC4 Variable key sized stream cipher; Often used in file
encryption and secure communications (SSL)
RC5 Variable block size and variable key length; uses 64-bit
block size; Fast, replacement for DES
RC6 Block cipher based on RC5, meets AES requirement
Chosen for speed and variable-key length capabilities
Designed mostly by Ronald Rivest
Each of the algorithms have different uses
Block Cipher
Transforms a fixed-length block of plain text into a block of
ciphertext
operate on a pre-determined block of bits (one byte, one word, 512
bytes, so forth), mixing key data in with the message data in a variety
of different ways
Common block ciphers:
DES and 3DES (in ECB and CBC mode)
Skipjack
Blowfish
RSA
AES
IDEA
SAFER
Stream Cipher
Use smaller units of plaintext than what are used with block
ciphers.
encrypts bits of the message at a time
typically bit-wise
They perform some operation (typically an exclusive OR) with one of
these key bits and one of the message bits
They either have a very long key (that eventually repeats) or a reusable
key that generates a repeatable but seemingly random string of bits.
Common stream ciphers:
RC4
DES and 3DES (running OFB or CFB mode)
SEAL
Block Cipher Modes
Defines how the block cipher algorithm is applied to the
data stream
Four Basic Modes
Electronic Code Book (ECB)
Cipher Block Chaining (CBC)
Cipher Feedback (CFB)
Output Feedback (OFB)
Electronic Codebook (ECB)
Ciphertext Block Chaining (CBC)
P
i
= D
k
(C
i
) C
i 1
, C
0
= IV
C
i
= E
k
(P
i
C
i 1
), C
0
= IV
Cipher Feedback (CFB)
C
i
= E
k
(C
i 1
) P
i
P
i
= E
k
(C
i 1
) C
i
C
o
= IV
Output Feedback (OFB)
Selecting a Block Cipher Mode
Small amounts of truly random data: ECB
Example: randomly generated keying material
Other modes can be used but ECB is most efficient
Protocols with crypto integrity protection: CBC, CFB, OFB
Arbitrary communications with arbitrary data: CBC, CFB
Repeated plaintext data is obscured
Constantly changing encryption keys defeat differential cryptanalysis
attacks
Asymmetric Key Algorithm
Also called public-key cryptography
Keep private key private
Anyone can see public key
separate keys for encryption and decryption (public and
private key pairs)
Examples:
RSA, DSA, Diffie-Hellman, ElGamal, PKCS
Asymmetric Encryption
Plaintext
ENCRYPTION
ALGORITHM
DECRYPTION
ALGORITHM
Ciphertext
Plaintext
Public Key Private Key
Asymmetric Key
Cryptography
Different keys
Asymmetric Key Algorithms
RSA the first and still most common implementation
DSA specified in NISTs Digital Signature Standard
(DSS), provides digital signature capability for
authentication of messages
Diffie-Hellman used for secret key exchange only, and not
for authentication or digital signature
ElGamal similar to Diffie-Hellman and used for key
exchange
PKCS set of interoperable standards and guidelines
Symmetric vs. Asymmetric Key
Symmetric Asymmetric
generally fast
Same key for both encryption and
decryption

Can be 1000 times slower
Uses two different keys (public and
private)
Decryption key cannot be calculated
from the encryption key
Key lengths: 512 to 4096 bits
Used in low-volume

Hash Functions
produces a condensed representation of a message (hashing)
The fixed-length output is called the hash or message digest
A hash function takes an input message of arbitrary length and
outputs fixed-length code.
Given x, we can compute the value f(x).
Given f(x), it is hard to get the value of x.
A form of signature that uniquely represents the data
Collision-free
Uses:
Verifying file integrity - if the hash changes, it means the data is either
compromised or altered in transit.
Digitally signing documents
Hashing passwords
Hash Functions
Message Digest (MD) Algorithm
Outputs a 128-bit fingerprint of an arbitrary-length input
MD4 is obsolete, MD5 is widely-used
Secure Hash Algorithm (SHA)
SHA-1 produces a 160-bit message digest similar to MD5
Widely-used on security applications (TLS, SSL, PGP, SSH, S/MIME,
IPsec)
SHA-256, SHA-384, SHA-512 are also commonly used, which can
produce hash values that are 256, 384, and 512-bits respectively
RIPEMD
Derived from MD4, but performs like SHA
RIPEMD-160 is the most popular version
Digital Signature
A digital signature is a message appended to a packet
The sender encrypts message with own private key instead
of encrypting with intended receivers public key
The receiver of the packet uses the senders public key to
verify the signature.
Used to prove the identity of the sender and the integrity of
the packet

Digital Signature
Two common public-key digital signature techniques:
RSA (Rivest, Shamir, Adelman)
DSS (Digital Signature Standard)
Used in a lot of things:
Email, software distribution, electronic funds transfer, etc
A common way to implement is to use a hashing algorithm
to get the message digest of the data, then use an
algorithm to sign the message
Digital Signature Process
1. Hash the data using one of the supported hashing
algorithms (MD5, SHA-1, SHA-256)
2. Encrypt the hashed data using the senders private key
3. Append the signature (and a copy of the senders public
key) to the end of the data that was signed)
DATA
HASH
(DATA)
DIGITAL
SIGNATURE
MD5/SHA-1 PRIVATE KEY
Signature Verification Process
1. Hash the original data using the same hashing algorithm
2. Decrypt the digital signature using the senders public key. All
digital signatures contain a copy of the signers public key
3. Compare the results of the hashing and the decryption. If the
values match then the signature is verified. If the values do not
match, then the data or signature was probably modified.
DATA
HASH
(DATA)
HASH
(DIGITAL SIG)
MD5/SHA-1
MATCH?
Message Authentication Code
Message authentication code provides
Integrity (checks that data has not been altered)
Authenticity (verifies the origin of data)
In the sender side, the message is passed through a MAC
algorithm to get a MAC (also called Tag)
In the receiver side, the message is passed through the same
algorithm. The output is compared with the received tag and
should match
Sender and receiver uses the same secret key
Hash-based Message Authentication Code (RFC2104)
Uses hash function to generate the MAC
HMACs are less affected by collisions than their underlying hashing
algorithms alone.
Questions
Public Key Infrastructure
Overview
Digital Certificates
Certificate Authority
Framework that builds the network of trust
Combines public key cryptography, digital signatures, to
ensure confidentiality, integrity, authentication,
nonrepudiation, and access control
Protects applications that require high level of security
Functions of a PKI
Registration
Initialization
Certification
Key pair recovery
Key generation
Key update
Cross-certification
Revocation
Source: http://commons.wikimedia.org
Components of a PKI
Certificate authority
The trusted third party
Trusted by both the owner of the certificate and the party relying upon
the certificate.
Validation authority
Registration authority
For big CAs, a separate RA might be necessary to take some work
off the CA
Identity verification and registration of the entity applying for a
certificate
Central directory

Certificates
Public key certificates bind public key values to subjects
A trusted certificate authority (CA) verifies the subjects identity
and digitally sign each certificate
Validates
Has a limited valid lifetime
Can be used using untrusted communications and can be
cached in unsecured storage
Because client can independently check the certificates signature
Certificate is NOT equal to signature
It is implemented using signature
Certificates are static
If there are changes, it has to be re-issued
Digital Certificate
Digital certificate basic element
of PKI; secure credential that
identifies the owner
Also called public key certificate
Certificate examples:
X509 (standard)
PGP (Pretty Good Privacy)

Every certificate contains
Body of the certificate
Version number, serial number, names of the issuer and subject
Public key associated with the subject
Expiration date (not before, not after)
Extensions for additional tributes
Signature algorithm
Used by the CA to sign the certificate
Signature
Created by applying the certificate body as input to a one-way hash
function. The output value is encrypted with the CAs private key to
form the signature value
Digital Certificate
To obtain a digital certificate, Alice must:
Make a certificate signing request to the CA
Alice sends to CA:
Her identifier IdA
Her public key KA_PUB
Additional information
CA returns Alices digital certificate, cryptographically
binding her identity to public key:
CertA = {IDA, KA_PUB, info, SigCA(IDA,KA_PUB,info)}
X.509
An ITU-T standard for a public key infrastructure (PKI) and
Privilege Management Infrastructure (PMI)
Assumes a strict hierarchical system of Certificate
Authorities (CAs)
RFC 1422 basis of X.509-based PKI
Current version X.509v3 provides a common baseline for
the Internet
Structure of a Certificate, certificate revocation (CRLs)
X.509 Certificate Usage
Fetch certificate
Fetch certificate revocation list
(CRL)
Check the certificate against the
CRL
Check signature using the
certificate

Certificate CRL
Signature
Check
Check
Certificate Authority
Issuer and signer of the certificate
Trusted (Third) Party
Based on trust model
Who to trust?
Types:
Enterprise CA
Individual CA (PGP)
Global CA (such as VeriSign)
Functions:
Enrolls and Validates Subscribers
Issues and Manages Certificates
Manages Revocation and Renewal of Certificates
Establishes Policies & Procedures

Certificate Revocation Lists
CA periodically publishes a data structure called a
certificate revocation list (CRL).
Described in X.509 standard.
Each revoked certificate is identified in a CRL by its serial
number.
CRL might be distributed by posting at known Web URL or
from CAs own X.500 directory entry.
Questions
IP Security (IPsec)
Overview
Introduction to VPN
IPsec Fundamentals
Tunnel and Transport Mode IPsec
Architecture and Components of IPsec
Internet Key Exchange
Configuring IPsec for IPv4 and IPv6
Virtual Private Network
Creates a secure tunnel over a public network
Client to firewall
Router to router
Firewall to firewall
Uses the Internet as the public backbone to access a
secure private network
Remote employees can access their office network
Two types:
Remote access
Site-to-site VPN
VPN Implementation
Hardware
Usually a VPN-type router
Pros: highest network throughput, plug and play, dual purpose
Cons: cost and lack of flexibility
Software
Ideal for two end-points in different organisations
Pros: flexible, and low relative cost
Cons: lack of efficiency, more labor training required, lower
productivity; higher labor costs
Firewall
Pros: cost effective, tri-purpose, hardens the operating system
Cons: still relatively costly

VPN Protocols
PPTP (Point-to-Point tunneling Protocol)
Developed by Microsoft to secure dial-up connections
Operates in the data-link layer
L2F (Layer 2 Forwarding Protocol)
Developed by Cisco
Similar as PPTP
L2TP (Layer 2 Tunneling Protocol)
IETF standard
Combines the functionality of PPTP and L2F
IPsec (Internet Protocol Security)
Open standard for VPN implementation
Operates on the network layer

Other Modern VPNs
MPLS VPN
Used for large and small enterprises
Pseudowire, VPLS, VPRN
GRE Tunnel
Packet encapsulation protocol developed by Cisco
Not encrypted
Implemented with IPsec
L2TP IPsec
Uses L2TP protocol
Usually implemented along with IPsec
IPsec provides the secure channel, while L2TP provides the tunnel
Advantages of VPN
Cheaper connection
Use the Internet connection instead of a private lease line
Scalability
Flexibility of growth
Efficiency with broadband technology
Availability
Available everywhere there is an Internet connection

Disadvantages of VPN
VPNs require an in-depth understanding of public network
security issues and proper deployment precautions
Availability and performance depends on factors largely
outside of their control
VPNs need to accommodate protocols other than IP and
existing internal network technology
IPsec
Provides Layer 3 security (RFC 2401)
Transparent to applications (no need for integrated IPsec support)
A set of protocols and algorithms used to secure IP data at
the network layer
Combines different components:
Security associations (SA)
Authentication headers (AH)
Encapsulating security payload (ESP)
Internet Key Exchange (IKE)
A security context for the VPN tunnel is established via the
ISAKMP
IPsec Standards
RFC 4301 The IP Security Architecture
Defines the original IPsec architecture and elements common to both AH
and ESP
RFC 4302
Defines authentication headers (AH)
RFC 4303
Defines the Encapsulating Security Payload (ESP)
RFC 2408
ISAKMP
RFC 5996
IKE v2 (Sept 2010)
RFC 4835
Cryptographic algorithm implementation for ESP and AH
Benefits of IPsec
Confidentiality
By encrypting data
Integrity
Routers at each end of a tunnel calculates the checksum or hash
value of the data
Authentication
Signatures and certificates
All these while still maintaining the ability to route through existing IP
networks

IPsec is designed to provide interoperable, high quality, cryptographically-
based security for IPv4 and IPv6 - (RFC 2401)
Benefits of IPsec
Data integrity and source authentication
Data signed by sender and signature is verified by the recipient
Modification of data can be detected by signature verification
Because signature is based on a shared secret, it gives source
authentication
Anti-replay protection
Optional; the sender must provide it but the recipient may ignore
Key management
IKE session negotiation and establishment
Sessions are rekeyed or deleted automatically
Secret keys are securely established and authenticated
Remote peer is authenticated through varying options

Different Layers of Encryption
Network Layer - IPsec
Link Layer Encryption
Application Layer SSL, PGP, SSH, HTTPS
IPsec Modes
Tunnel Mode
Entire IP packet is encrypted and becomes the data component of a
new (and larger) IP packet.
Frequently used in an IPsec site-to-site VPN
Transport Mode
IPsec header is inserted into the IP packet
No new packet is created
Works well in networks where increasing a packets size could cause
an issue
Frequently used for remote-access VPNs
Tunnel vs. Transport Mode IPsec
Payload
TCP
Header
IP
Header
Without IPsec
Transport Mode
IPsec
Tunnel Mode
IPsec
Payload
TCP
Header
IP
Header
IPsec
Header
IP
Header
Payload
TCP
Header
IP
Header
IPsec
Header
New IP
Header
IPsec Architecture
ESP
AH
IKE
IPsec Security Policy
Encapsulating Security
Payload
Authentication Header
The Internet Key Exchange
Security Associations (SA)
A collection of parameters required to establish a secure
session
Uniquely identified by three parameters consisting of
Security Parameter Index (SPI)
IP destination address
Security protocol (AH or ESP) identifier
An SA is unidirectional
Two SAs required for a bidirectional communication
A single SA can be used for AH or ESP, but not both
must create two (or more) SAs for each direction if using both AH and
ESP

Security Associations
INTERNET
A Security Association (SA) Defines:
- communicating parties [aka selectors]
- security services used to protect traffic [AH / ESP]
- cryptographic algorithm used
- cipher key length
- lifetime of key

A unique 32-bit identification number that is part of the
Security Association (SA)
It enables the receiving system to select the SA under
which a received packet will be processed.
Has only local significance, defined by the creator of the
SA.
Carried in the ESP or AH header
When an ESP/AH packet is received, the SPI is used to
look up all of the crypto parameters

How to Set Up SA
Manually
Sometimes referred to as manual keying
You configure on each node:
Participating nodes (I.e. traffic selectors)
AH and/or ESP [tunnel or transport]
Cryptographic algorithm and key
Automatically
Using IKE (Internet Key Exchange)
ISAKMP
Internet Security Association and Key Management
Protocol
Used for establishing Security Associations (SA) and
cryptographic keys
Only provides the framework for authentication and key
exchange, but key exchange is independent
Key exchange protocols
Internet Key Exchange (IKE) and Kerberized Internet Negotiation of
Keys (KINK)
Selectors
Defines when to create an SA and what the SA will be used
for
Classifies the type of traffic requiring IPsec protection and
the kind of protection to be applied.
Elements of a selector:
Source IP address
Destination IP address
Protocol (TCP or UDP)
Upper layer protocol
Example: use ESP with NULL encryption and HMAC-SHA1 for routing updates, but
use ESP with 3DES and SHA-1 for telnet and TFTP access for a router
Authentication Header (AH)
Provides source authentication and data integrity
Protection against source spoofing and replay attacks
Authentication is applied to the entire packet, with the
mutable fields in the IP header zeroed out
If both AH and ESP are applied to a packet, AH follows ESP
Operates on top of IP using protocol 51
In IPv4, AH protects the payload and all header fields
except mutable fields and IP options (such as IPsec option)

AH Header Format
Next Header (8 bits): indicates which upper layer protocol is protected (UDP, TCP, ESP)
Payload Length (8 bits): size of AH in 32-bit longwords, minus 2
Reserved (16 bits): for future use; must be set to all zeroes for now
SPI (32 bits): arbitrary 32-bit number that specifies to the receiving device which security association is being
used (security protocols, algorithms, keys, times, addresses, etc)
Sequence Number (32 bits): start at 1 and must never repeat. It is always set but receiver may choose to
ignore this field
Authentication Data: ICV is a digital signature over the packet and it varies in length depending on the
algorithm used (SHA-1, MD5)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Next Header Payload Length Reserved
Sequence Number
Authentication Data
[ Integrity Check Value (ICV) ]
Encapsulating Security Payload (ESP)
Uses IP protocol 50
Provides all that is offered by AH, plus data confidentiality
It uses symmetric key encryption
Must encrypt and/or authenticate in each packet
Encryption occurs before authentication
Authentication is applied to data in the IPsec header as well
as the data contained as payload

ESP Header Format
SPI: arbitrary 32-bit number that specifies SA to the receiving device
Seq #: start at 1 and must never repeat; receiver may choose to ignore
IV: used to initialize CBC mode of an encryption algorithm
Payload Data: encrypted IP header, TCP or UDP header and data
Padding: used for encryption algorithms which operate in CBC mode
Padding Length: number of bytes added to the data stream (may be 0)
Next Header: the type of protocol from the original header which appears in the encrypted part of the
packet
Authentication Header: ICV is a digital signature over the packet and it varies in length depending on
the algorithm used (SHA-1, MD5)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Next Header Padding Length
Payload Data (Variable)
Padding (0-255 bytes)
Initialization Vector (IV)
Sequence Number
Authentication Data (ICV)
E
N
C
R
Y
P
T
E
D

Packet Format Alteration for AH
Transport Mode
Original
IP Header TCP/UDP Data
Original
IP Header
AH
Header TCP/UDP Data
Without AH
With AH
Authenticated except for
mutable fields in IP header
ToS
TTL
Header Checksum
Offset
Flags
Packet Format Alteration for ESP
Transport Mode
Original
IP Header
TCP/UDP Data
Original
IP Header
ESP
Header
Encapsulating Security Payload
Before applying
ESP:
After applying
ESP:
Encrypted
ESP
Authentication
Authenticated
TCP/UDP Data
ESP
Trailer
Packet Format Alteration for AH
Tunnel Mode
Original
IP Header
TCP/UDP Data
New
IP Header
AH
Header
Data
Before applying
AH:
After applying
AH:
Authenticated except for
mutable fields in new IP header
Original
IP Header
ToS
TTL
Header Checksum
Offset
Flags
Packet Format Alteration for ESP
Tunnel Mode
Original
IP Header
TCP/UDP Data
New
IP Header
ESP
Header
Encapsulating Security Payload
Before applying
ESP:
After applying
ESP:
Encrypted
ESP
Authentication
Authenticated
Original
IP Header
TCP/UDP Data
ESP
Trailer
An IPsec component used for performing mutual
authentication and establishing and maintaining Security
Associations. (RFC 5996)
Typically used for establishing IPsec sessions
A key exchange mechanism
Five variations of an IKE negotiation:
Two modes (aggressive and main modes)
Three authentication methods (pre-shared, public key encryption,
and public key signature)
Uses UDP port 500

IKE Modes
Mode Description
Main mode Three exchanges of information between IPsec peers.
Initiator sends one or more proposals to the other peer
(responder)
Responder selects a proposal
Aggressive Mode Achieves same result as main mode using only 3 packets
First packet sent by initiator containing all info to establish
SA
Second packet by responder with all security parameters
selected
Third packet finalizes authentication of the ISAKMP
session
Quick Mode Negotiates the parameters for the IPsec session.
Entire negotiation occurs within the protection of ISAKMP
session
Phase I
Establish a secure channel (ISAKMP SA)
Using either main mode or aggressive mode
Authenticate computer identity using certificates or pre-shared secret

Phase II
Establishes a secure channel between computers intended for the
transmission of data (IPsec SA)
Using quick mode

Overview of IKE
Traffic which needs
to be protected
IPsec Peer
IPsec Peer
IKE Phase 1
Secure communication channel
IKE Phase 2
IPsec Tunnel
Secured traffic exchange
1
2
3
4
ISAKMP Header Format
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Initiator Cookie
Total Length of Message
Flags
Responder Cookie
Next Payload Exchange Type
Message ID
Major
Version
Minor
Version
ISAKMP Message Format
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Next Payload: 1byte; identifier for next payload in message. If it is the last payload
It will be set to 0

Reserved: 1byte; set to 0

Payload Length: 2 bytes; length of payload (in bytes) including the header

Payload: The actual payload data

Next Payload Reserved Payload Length
Payload
Next Payload Reserved Payload Length
Payload
ISAKMP HEADER
IKE Phase 1 (Main Mode)
Main mode negotiates an ISAKMP SA which will be used to
create IPsec SAs
Three steps
SA negotiation (encryption algorithm, hash algorithm, authentication
method, which DF group to use)
Do a Diffie-Hellman exchange
Provide authentication information
Authenticate the peer

IKE Phase 1 (Main Mode)
Responder
Initiator
1
2
IKE Message 1 (SA proposal)
IKE Message 2 (accepted SA)
IKE Message 3 (DH public value, nonce)
IKE Message 4 (DH public value, nonce)
IKE Message 5 (Authentication material, ID)
IKE Message 6 (Authentication material, ID)
4
3
Negotiate
IKE Policy
Authenticated
DH Exchange
Compute DH shared secret
and derive keying material
Protect IKE
Peer Identity
Internet
(Encrypted)
IKE Phase 1 (Aggressive Mode)
Uses 3 (vs 6) messages to establish IKE SA
No denial of service protection
Does not have identity protection
Optional exchange and not widely implemented

IKE Phase 2 (Quick Mode)
All traffic is encrypted using the ISAKMP Security
Association
Each quick mode negotiation results in two IPsec Security
Associations (one inbound, one outbound)
Creates/refreshes keys

IKE Phase 2 (Quick Mode)
Responder
Initiator
3
Compute keying material
Internet
Message 1 (authentication/keying material and SA proposal)
Message 2 (authentication/keying material and accepted SA)
Message 3 (hash for proof of integrity/authentication)
1
2
5
Validate
message 1
7
4
6
Validate
message 3
Validate
message 2
IPv6 and IPsec Standards
IPv6 node requirements draft
ESP and AH must be supported
Must support manual configuration of SA
Key management SHOULD be supported (IKEv1, IKEv2, Kerberos,
etc)
IKEv1 vs IKEv2
IKEv2 incorporates extensions made to IKEv1
IKEv2 has support for multiple addresses for traffic selectors which is
useful in Mobile scenarios
Are vendors supporting IKEv2?
IPv6 AH Transport Mode
Original
IP Header
Original
IP Header
Hop-by-hop, DST Options*,
Routing, Fragment
Extension
Headers
TCP
DATA AH TCP
DATA
Mutable field processing Immutable fields
Authenticated except for mutable fields
* DST options extension header could appear before, after, or
both before and after the AH header.
DST
Options*
Mutable Fields:
- DSCP
- ECN
- Flow Label
- Hop Limit
IPv6 AH Tunnel Mode
Original
IP Header
Original
IP Header
Hop-by-hop, Dest*,
Routing, Fragment
Extension
Headers
TCP
DATA AH TCP
DATA
Mutable field processing
Immutable fields
Authenticated except for mutable fields
New
IP Header
Hop-by-hop, Dest*,
Routing, Fragment
Mixed inner and outer IP versions are allowed (I.e. IPv4 over IPv6 or
IPv6 over IPv4)
Mutable Fields:
- DSCP
- ECN
- Flow Label
- Hop Limit
IPv6 IPsec AH Considerations
INTERNET
Route authentication:
OSPFv3 requires IPsec
Replace MD-5 for other routing protocols?

Most initial host implementations support AH

Is AH going to be deprecated?
Original
IP Header
Original
IP Header
Hop-by-hop, DST Options*,
Routing, Fragment
Extension
Headers
TCP
DATA ESP TCP
DATA
Encryption
Integrity
ESP
Trailer
ESP
ICV
IPv6 ESP Transport Mode
DST
Options*
* DST options extension header could appear before, after, or
both before and after the ESP header. Since ESP protects
only fields after the ESP header, it generally will be desirable
to place the destination options header(s) after the ESP
header.
IPv6 ESP Tunnel Mode
New
IP Header
Original
IP Header
New Ext
Headers*
Extension
Headers
TCP
DATA ESP TCP
DATA
Encryption
Integrity
ESP
Trailer
ESP
ICV
Original Ext
Headers*
Original
IP Header
IPv6 IPsec ESP
INTERNET
- ESP with NULL encryption can be used for router authentication

- National /Corporate policies will dictate where encryption is feasible

- Distributed firewalls with IPsec hooks can still give stateful inspection
IPsec Best Practices
Use IPsec to provide integrity in addition to encryption
Use ESP option
Use strong encryption algorithms
AES instead of DES
Use a good hashing algorithm
SHA instead of MD5
Reduce the lifetime of the Security Association (SA) by
enabling Perfect Forward Secrecy (PFS)
Increases processor burden so do this only if data is highly sensitive
Configuring IPsec
Step 1: Configure the IKE Phase 1 Policy (ISAKMP Policy)
crypto isakmp policy [priority]
Step 2: Set the ISAKMP Identity
crypto isakmp identity {ipaddress|hostname}
Step 3: Configure the IPsec transfer set
crypto ipsec transform-set transform-set-name
<transform1> <transform2> mode [tunnel|transport]
crypto ipsec security-association lifetime seconds
seconds
Configuring IPsec
Step 5: Creating map with name
Crypto map crypto-map-name seq-num ipsec-isakmp
Match address access-list-id
Set peer [ipaddress|hostname]
Set transform-set transform-set-name
Set security-association lifetime seconds seconds
Set pfs [group1|group2]
Step 6: Apply the IPsec Policy to an Interface
Crypto map crypto-map-name local-address interface-
id

IPsec Layout
R1 R2
Encrypted session
Public Network
Router Configuration
crypto isakmp policy 1
authentication pre-share
encryption aes
hash sha
group 5
crypto isakmp key Training123 address 172.16.11.66
!
crypto ipsec transform-set ESP-AES-SHA esp-aes esp-sha-hmac
!
crypto map LAB-VPN 10 ipsec-isakmp
match address 101
set transform-set ESP-AES-SHA
set peer 172.16.11.66

Phase 1 SA
Encryption and
authentication
Phase 2 SA
Router Configuration
int fa 0/1
crypto map LAB-VPN
Exit
!
access-list 101 permit ip 172.16.16.0
0.0.0.255 172.16.20.0 0.0.0.255

Apply to an
outbound interface
Define interesting
VPN traffic
IPsec Debug Commands
sh crypto ipsec sa
sh crypto isakmp peers
sh crypto isakmp sa
sh crypto map
Questions
As part of your membership agreement with APNIC, all
Members are required to register their resources in the
APNIC database.
First allocation/assignment, APNIC will create:
Inetnum or inet6num object
Autnum object (if you received an ASN)
Maintainer object (to protect your data)
Role object
Members must keep records up to date:
Whenever there is a change in contacts
When new resources are received
When resources are sub-allocated or assigned

What is the APNIC Database?
Public network management database
Operated by Internet Registries
Public data only
(For private data, please see Privacy of customer assignment module)
Tracks network resources
IP addresses, ASNs, Reverse Domains, Routing policies
Records administrative information
Contact information (persons/roles)
Authorization
Whois Database Query - Clients
Standard whois client
Included with many Unix distributions
RIPE extended whois client
http://ftp.apnic.net/apnic/dbase/tools/ripe-dbase-client.tar.gz
Query via the APNIC website
http://www.apnic.net/apnic-bin/whois2.pl
Query clients MS Windows etc

Object Types
OBJECT PURPOSE
person contact persons
role contact groups/roles
inetnum IPv4 addresses
Inet6num IPv6 addresses
aut-num Autonomous System number
domain reverse domains
route prefixes being announced
mntner (maintainer) data protection
mnt-irt Incident Response Team
http://www.apnic.net/db/
Database Object
An object is a set of attributes and values
Each attribute of an object
Has a value
Has a specific syntax
Is mandatory or optional
Is single or multi-valued
Some attributes are
Primary (unique) keys
Lookup keys for queries
Inverse keys for queries
Object templates illustrate this structure
Inter-Related Objects
inetnum:
202.64.10.0 202.64.10.255

admin-c: EC196-AP
tech-c: ZU3-AP

mnt-by: MAINT-WF-EX

IPv4 addresses
person/role:

nic-hdl: ZU3-
AP

Contact info
mntner:
MAINT-WF-EX

Data protection
person/role:

nic-hdl: EC196-
AP

Contact info
New Members
If you are receiving your first allocation or assignment,
APNIC will create the following objects for you:
role object
inetnum or inet6num object
maintainer object (to protect your data)
aut-num object (if you received an ASN)
Information is taken from your application for resources and
membership
Inetnum / Inet6num Objects
Contains IP allocation and assignment information
APNIC creates an inetnum (or inet6num) object for each
allocation or assignment they make to the Member
All members must create inetnum (or inet6num) objects for
each sub-allocation or assignment they make to customers
Whois Inet6num Example
Person Object
Represents a contact person for an organization
Every Member must have at least one contact person registered
Large organizations often have several contacts for different
purposes
Is referenced in other objects
Has a nic-hdl
Eg. EC17-AP
What is a nic-hdl?
Unique identifier for a person or role
Represents a person or role object
Referenced in objects for contact details
(inetnum, aut-num, domain)
format: <XXXX-AP>
Eg: EC196-AP
Person: Eric Chu
address: ExampleNet Service Provider
address: Level 1 33 Park Road Milton
address: Wallis and Futuna Islands
country: WF
phone: +680-368-0844
fax-no: +680-367-1797
e-mail: [email protected]
nic-hdl: EC196-AP
mnt-by: MAINT-WF-EX
changed: [email protected] 20020731
source: APNIC
Role Object
Represents a group of contact persons for an organization
Eases administration
Can be referenced in other objects instead of the person objects for
individuals
Also has a nic-hdl
Eg. HM20-AP

NOC Role Admin Role
How a Role Object Works
Role Object is used instead of a Person Object as a
reference in other objects
If a contact leaves the organization:
New Person Object is created
The nic-hdl of the new contact replaces nic-hdl of the old person in
the Role Object
Old Person Object is deleted
This means only a single replacement is required instead of
many

Replacing Contacts in the DB
- Using Person Objects
E. Chu is leaving my organization.
A. Nagali is replacing him.

1. Create a Person Object for new
contact (E. Chu)
2. Find all objects containing old
contact (E. Chu)
3. Update all objects, replacing
old contact (EC196-AP) with
new contact (AN3-AP)
4. Delete old contacts (EC196-
AP) Person Object
inetnum:
202.0.10.0

person:

EC196-AP
inetnum:
202.0.15.192

inetnum:
202.0.12.127

person:

AN3-AP
AN3-AP
AN3-AP
AN3-AP
Replacing Contacts in the DB
Using a Role Object
E. Chu is leaving my organization.
A. Nagali is replacing him.

1. Create a Person Object for new
contact (A. Nagali)
2. Replace old contact (EC196-
AP) with new contact (AN3-AP)
in Role Object
3. Delete old contacts Person
Object.

inetnum:
202.0.10.0

EIPA91-AP
Person:

EC196-AP
inetnum:
202.0.15.192

EIPA91-AP
inetnum:
202.0.12.127

EIPA91-AP
role:

EIPA-91-AP
KX17-AP
AB1-AP
CD2-AP

AN3-AP
person:

AN3-AP
No need to update any other objects!
My Role Object contains all contact info,
that is referenced in all my objects.
Whois - Role vs Person Objects
IRT Object
Incident Response Team (IRT)
Dedicated abuse handling teams (not netops)
Implemented in Nov 2010 through Prop-079
Abuse contact information
Mandatory object reference in inetnum, inet6num, and aut-
num objects
IRT Object
Why provide abuse contact
Dedicated contacts or team that specifically resolve computer
security incidents
Efficient and accurate response
Stops the tech-c and admin-c from getting abuse reports
Shared response to address abuse

Database Protection - Maintainers
protects other objects in the APNIC Whois Database
used to prevent unauthorized persons from changing the
details in whois
Multiple levels of maintainers exist in a hierarchical manner
Maint-by
Maint-lower
Applied to any object created directly below that maintainer
object

Database Protection
Authorization
mnt-by references as maintainer object
Can be found in all database objects
mnt-by should be used with every object
Authentication
Updates to an object must pass the authentication rule specified by
its maintainer
Authentication methods (using auth attribute)
Crypt-PW
PGP GNUPG
MD5
Database Protection
Maintainer Object
Maintainer Hierarchy Diagram
Allocated to APNIC
Maint-by can only be changed by
IANA
Allocated to Member
APNIC
Sub-allocated to Customer
Members
Customer Privacy
Privacy issues
Concerns about publication of customer information
Increasing government concern
APNIC legal risk
Legal responsibility for accuracy and advice
Damages incurred by maintaining inaccurate personal data
Customer data is hard to maintain
Customer assignment registration is still mandatory
What Needs to be Visible?
Questions
Resource Public Key
Infrastructure (RPKI)
SIDR Working Group
Secure Inter-Domain Routing (SIDR)
Its purpose is to reduce vulnerabilities to the inter-domain
routing system
Addresses two vulnerabilities:
Is an Autonomous System authorized to originate an IP prefix?
Is the AS-Path represented in the route the same as the path through
which the NLRI traveled?
Projects:
PKI, RPKI, BGPsec
Source: SIDR WG
https://datatracker.ietf.org/wg/sidr/charter/
BGP Security (BGPsec)
Extension to BGP that provides improved security for BGP
routing
Currently an IETF Internet draft
Implemented via a new optional non-transitive BGP path
attribute that contains a digital signature
Two things:
BGP Prefix Origin Validation (using RPKI)
BGP Path Validation
Three Pieces
RPKI Resource Public Key Infrastructure, the Certificate
Infrastructure to Support the other Pieces (deployed at all
RIRs)
Origin Validation Using the RPKI to detect and prevent
mis-originations of someone elses prefixes (in deployment)
AS-Path Validation AKA BGPsec Prevent Path Attacks on
BGP (future work)
What is RPKI?
Resource Public Key Infrastructure (RPKI)
A robust security framework for verifying the association between
resource holder and their Internet resources
Created to address the issues in RFC 4593
Uses X.509 v3 certificates
With RFC3779 extensions
Helps to secure Internet routing by validating routes
Proof that prefix announcements are coming from the legitimate holder
of the resource
A system to manage the creation and storage of digital
certificates and the associated Route Origin Authorization
documents

RFCs on RPKI
RFC 6810 The Resource Public Key Infrastructure (RPKI)
to Router Protocol (January 2013) - Standard
RFC 6480 An Infrastructure to Support Secure Internet
Routing (Feb 2012) - informational
RFC 6481 A Profile for Resource Certificate Repository
Structure (Feb 2012) - standard
RFC 6491 RPKI Objects Issued by IANA
RFC 6493 The RPKI Ghostbusters Record
RFC 6487 A Profile for X.509 PKIX Resource Certificate
A bit of History
1986 Bellovin & Perlman identify the vulnerability in DNS
and Routing
1999 - National Academies study called it out
2000 S-BGP X.509 PKI to support Secure BGP - Kent,
Lynn, et al.
2003 NANOG S-BGP Workshop
2006 RPKI.NET(for ARIN) & APNIC start work on RPKI.
RIPE starts in 2008.
2009 RPKI.NET Open Testbed and running code in test
routers
Internet Routing
The Internet
Traffic
202.12.29.0/24
Announce
202.12.29.0/24
Global Routing Table
4.128/9
60.100/16
60.100.0/20
135.22/16

Global Routing Table
4.128/9
60.100/16
60.100.0/20
135.22/16
202.12.29.0/24

202.12.29.0/24
Autonomous System (AS)
Collection of networks with same routing policy
Single routing protocol
Usually under single ownership, trust and administrative
control
AS 100
What is AS path?
An AS path example:

Benefits of RPKI - Routing
Prevents Route Hijacking
when an entity participating in Internet routing announces a prefix
without authorization
Reason: malicious attack
Prevents mis-origination
A prefix that is originated by an AS which does not own it
Reason: configuration mistake

Right to Resources
ISP gets their resources from the RIR
ISP notifies its upstream of the prefixes to be announced
Upstream _MUST_ check the Whois database if resource
has been delegated to customer ISP.
We need to be able to Authoritatively prove who owns an IP Prefix and
what AS(s) may announce it.

X.509 Certificate
Resource certificates are based on the X.509 certificate
format - RFC 5280
Extended by RFC 3779 this extension binds a list of
resources (IP, ASN) to the subject of the certificate
X.509 Certificate with 3779 Extension
SIA Subject Information Access;
contains a URI that references
the directory X.509 Certificate
RFC 3779
Extension
SIA
Owner's Public Key
Two Components
Certificate Authority (CA)
Internet Registries (RIR, NIR, Large LIR)
Issue certificates for customers
Allow customers to use the CAs GUI to issue ROAs for their prefixes
Relying Party (RP)
Software which gathers data from CAs
98.128.0.0/16
Public Key
98.128.0.0/20
Public Key
98.128.16.0/20
Public Key
98.128.32.0/19
Public Key
98.128.16.0/24
Public Key
98.128.17.0/24
Public Key
Cert/RGnet
Cert/Rob Cert/Randy
Cert/ISC
Cert/PSGnet
Cert/ARIN
CA
CA
CA CA
CA CA
Certificate
Hierarchy
follows
Allocation
Hierarchy
S
I
A

router bgp 651nn

bgp rpki server tcp 10.0.0.3 port 43779 refresh 60
bgp rpki server tcp 147.28.0.84 port 93920 refresh 60

Configuring Routers to Get ROAs
r0.sea#show bgp 192.158.248.0/24
BGP routing table entry for 192.158.248.0/24, version 3043542
Paths: (3 available, best #1, table default)
6939 27318
206.81.80.40 (metric 1) from 147.28.7.2 (147.28.7.2)
Origin IGP, metric 319, localpref 100, valid, internal,
best
Community: 3130:391
path 0F6D8B74 RPKI State valid
2914 4459 27318
199.238.113.9 from 199.238.113.9 (129.250.0.19)
Origin IGP, metric 43, localpref 100, valid, external
Community: 2914:410 2914:1005 2914:3000 3130:380
path 09AF35CC RPKI State valid

Valid
r0.sea#show bgp 198.180.150.0
Advertised to update-groups:
2 5 6 8
Refresh Epoch 1
1239 3927
144.232.9.61 (metric 11) from 147.28.7.2 (147.28.7.2)
Origin IGP, metric 759, localpref 100, valid, internal
Community: 3130:370
path 1312CA90 RPKI State invalid
Invalid
r0.sea#show bgp 64.9.224.0
Advertised to update-groups:
2 5 6
Refresh Epoch 1
1239 3356 36492
144.232.9.61 (metric 11) from 147.28.7.2 (147.28.7.2)
Origin IGP, metric 4, localpref 100, valid, internal
Community: 3130:370
path 11861AA4 RPKI State not found
Not Found
Validated ROA Payload (VRP)
A Prefix is Covered by a VRP when the VRP
prefix length is less than or equal to the Route
prefix length
98.128.0.0/16
98.128.0.0/12-16
98.128.0.0/16-24
98.128.0.0/20-24
Covers
Covers
No. Its Longer
BGP
VRP
VRP
VRP
Validated ROA Payload (VRP)
Prefix is Matched by a VRP when the Prefix is Covered by
that VRP , prefix length is less than or equal to the VRP
max-len, and the Route Origin AS is equal to the VRPs AS
98.128.0.0/16 AS 42
98.128.0.0/12-16 AS 42
98.128.0.0/16-24 AS 666
98.128.0.0/20-24 AS 42
Matched
No. AS Mismatch
No. VRP Longer
BGP
VRP
VRP
VRP
Matching and Validity
98.128.0.0/16-24 AS 6
BGP
VRP0
98.128.0.0/16-20 AS 42
VRP1
98.128.0.0/12 AS 42 NotFound, shorter than VRPs
BGP
98.128.0.0/16 AS 42 Valid, Matches VRP1
BGP
BGP
98.128.0.0/24 AS 42 Invalid, longer than VRP with AS 42
BGP
Policy - Fairly Secure
route-map validity-0
match rpki valid
set local-preference 100
match rpki not-found
! invalid is dropped
Policy - Paranoid
match rpki valid
! everything else dropped
Policy - Security Geek
match rpki invalid
! everything else dropped
After AS-Path
match rpki not-found
set metric 100
match rpki invalid
set metric 150
set metric 50
Set a Community
match rpki valid
set community 3130:400
match rpki invalid
RPKI at the Registries
RIPE seriously deployed with a few thousand LIRs and
thousands of ROAs
APNIC is operational, 550 LIRs, and moving forward,
moving to RIPEs GUI
ARIN is doing their best to make RPKI deployment very
hard
LACNIC is deployed and has O(100) LIRs
AFRINIC is deployed with O(25) LIRs
Per-RIR Statistics
Half are Two LIRs
Much Better
Than IPv6
Embarrassing
Router Origin Validation
Cisco IOS solid in 15.2
Cisco IOS/XR shipped in 4.3.2
Juniper shipped in 12.2
AlcaLu in development
RPKI Implementations
RIPE/NCC CA (partial closed) & RP (partial open)
APNIC CA only Closed Source
RTRlib/Berlin RP only Open Source
BBN RP Only Open Source
RPKI.NET CA & RP Open Source
RPKI-related Commands
r0.sea#show ip bgp rpki ?
servers Display RPKI cache server
information
table Display RPKI table entries

Check Server
r0.sea#show ip bgp rpki servers
BGP SOVC neighbor is 10.0.0.3/43779 connected to port 43779
Flags 0, Refresh time is 600, Serial number is 1304239609
InQ has 0 messages, OutQ has 0 messages, formatted msg 345
Session IO flags 3, Session flags 4008
Neighbor Statistics:
Nets Processed 624

Connection state is ESTAB, I/O status: 1, unread input bytes: 0
Connection is ECN Disabled
Mininum incoming TTL 0, Outgoing TTL 255
Local host: 199.238.113.10, Local port: 57932
Foreign host: 10.0.0.3, Foreign port: 43779
Connection tableid (VRF): 0
Look at Table
router1#show ip bgp rpki table
76 BGP sovc network entries using 6688 bytes of memory
78 BGP sovc record entries using 1560 bytes of memory
Network Maxlen Origin-AS Source Neighbor
98.128.0.0/24 24 3130 0 10.0.0.3/43779
98.128.0.0/16 16 3130 0 10.0.0.3/43779
98.128.6.0/24 24 4128 0 10.0.0.3/43779
98.128.9.0/24 24 3130 0 10.0.0.3/43779
98.128.30.0/24 24 1234 0 10.0.0.3/43779
128.224.1.0/24 24 3130 0 10.0.0.3/43779
129.6.0.0/17 17 49 0 10.0.0.3/43779
129.6.112.0/24 24 10866 0 10.0.0.3/43779
129.6.128.0/17 17 49 0 10.0.0.3/43779
147.28.0.0/16 16 3130 0 10.0.0.3/43779
Look at BGP Table
r0.sea#sh ip bgp
Network Next Hop Metric LocPrf Weight
Path
* i I198.180.150.0 144.232.9.61 100 0 1239 3927 i
*> I 199.238.113.9 0 2914 3927 i
* I 129.250.11.41 0 2914 3927 i
*> V198.180.152.0 199.238.113.9 0 2914 4128 i
* V 129.250.11.41 0 2914 4128 i
*> N198.180.155.0 199.238.113.9 0 2914 22773 i
* N 129.250.11.41 0 2914 22773 i
*> N198.180.160.0 199.238.113.9 0 2914 23308 13408
5752 i
* N 129.250.11.41 0 2914 23308 13408
5752 i
Resource Certification
Resource Certification
RIRs have been developing a new service for their
members
APNIC has now launched Resource Certification for the AP
region
The goal is to improve the security of inter-domain routing
and augmenting the information published in the APNIC
Whois Database
Resource Certification Benefits
Routing information corresponds to properly delegated
address resources
Resource Certification gives resource holders proof that
they hold certain resources
Resource holders can attest to those resources when
distributing them
Resource Certification is a highly robust means of
preventing the injection of false information into the
Internet's routing system.
APNIC Resource Certification
A robust security framework for verifying the association
between resource holders and their Internet resources.
Initiative from APNIC aimed at
improving the security of inter-domain routing, and
augmenting the information published in the Whois database
Verifies a holders current right-of-use over an Internet
resource

How it Works
Resource Certification (APNIC)
Verify signed data using the signers public key
Verify public key through a chain of interlocking certificates
that connect a Trust Anchor to the signers public key
certificate.
This is what we refer to as RPKI
Why its important:
Routing advertisements is now verifiable
Creating ROA Records
Login to MyAPNIC, then Resources -> Certification
Adding ROA Records
Simple view and add using the form
Deleting ROA Records
RPKI Validation
RPKI-capable routers can fetch the validated ROA dataset
from a trust anchor
BGP states: VALID, INVALID, UNKNOWN
Valid a matching VRP* was found
Invalid
Unknown

Questions
Route Filtering
Route Filtering
Types of prefixes in IP
core network:
Internal Prefixes
External prefixes
Downstream customers
Internet prefixes
Route Filtering
Internal prefixes originated
in IP core network
Loopback
Transport
Connect inter-regional networks
Point-to-point
Infrastructure point-to-point
Customer side point-to-point
Data centre
Some ISP originate from separate
AS if it is a large public hosting
operation and multihome DC
Route Filtering
Loopback Prefix
Prefix size /128
Advertised in IGP i.e. OSPF
Scope within IP core network
Can be summarize in IGP i.e. OSPF if the number of loopback prefixes are large
within the region
Route Filtering
Router1#sh ipv6 route
IPv6 Routing Table - default - 51 entries
Codes: C - Connected, L - Local, S - Static, U - Per-user Static route
B - BGP, HA - Home Agent, MR - Mobile Router, R - RIP
I1 - ISIS L1, I2 - ISIS L2, IA - ISIS interarea, IS - ISIS summary
D - EIGRP, EX - EIGRP external, ND - Neighbor Discovery, l - LISP
O - OSPF Intra, OI - OSPF Inter, OE1 - OSPF ext 1, OE2 - OSPF ext 2
ON1 - OSPF NSSA ext 1, ON2 - OSPF NSSA ext 2

LC 2406:6400::1/128 [0/0]
via Loopback0, receive
OI 2406:6400::2/128 [110/10]
via FE80::C801:1FF:FEAE:1C, Ethernet1/0
O 2406:6400::3/128 [110/10]
via FE80::C802:1FF:FEAE:1D, Ethernet1/1
OI 2406:6400::4/128 [110/21]
OI 2406:6400::5/128 [110/11]
OI 2406:6400::6/128 [110/21]
OI 2406:6400::7/128 [110/21]
OI 2406:6400::8/128 [110/11]
OI 2406:6400::9/128 [110/21]
OI 2406:6400::10/128 [110/21]
OI 2406:6400::11/128 [110/11]
OI 2406:6400::12/128 [110/21]
Loopback prefixes in
Training ISP network:
Route Filtering
Transport Prefix
Prefix size can be /64~/48
/48 is preferred if BGP traffic engineering required in future
Route Filtering
O - OSPF Intra, OI - OSPF Inter, OE1 - OSPF ext 1, OE2 - OSPF ext
2

OI 2406:6400:2::/48 [110/11]
OI 2406:6400:3::/48 [110/11]
Transport prefixes in Training
ISP network:
Route Filtering
Prefixes advertised/originated in IP core network
Point-to-point
Infrastructure point-to-point
Prefix size /64 [/127 on interface configuration according to rfc-6164]
Can be summarize in IGP i.e. OSPF if the number of p-to-p prefixes are
large within the region
Customer side point-to-point
Prefix size /64 [/127 on interface configuration according to rfc-6164]
Advertise from EGP i.e. iBGP (Not OSPF)
Summarization in iBGP using network statement and pull up route [Atomic
summarization]
Route Filtering
O - OSPF Intra, OI - OSPF Inter, OE1 - OSPF ext 1, OE2 - OSPF ext 2

C 2406:6400:E::/64 [0/0]
via Ethernet1/0, directly connected
O 2406:6400:E:1::/64 [110/20]
via FE80::C802:1FF:FEAE:1D, Ethernet1/1
C 2406:6400:E:2::/64 [0/0]
via Ethernet1/1, directly connected
OI 2406:6400:E:10::/64 [110/21]
OI 2406:6400:E:11::/64 [110/21]
OI 2406:6400:E:12::/64 [110/31]
OI 2406:6400:E:20::/64 [110/21]
OI 2406:6400:E:21::/64 [110/21]
OI 2406:6400:E:22::/64 [110/31]
OI 2406:6400:E:30::/64 [110/21]
OI 2406:6400:E:31::/64 [110/21]
OI 2406:6400:E:32::/64 [110/31]
Infrastructure p-to-p prefixes
in Training ISP network:
Route Filtering
2

S 2406:6400:10::/48 [1/0]
via Null0, directly connected
B 2406:6400:14::/48 [200/0]
via 2406:6400::3
B 2406:6400:18::/48 [200/0]
via 2406:6400::4
B 2406:6400:1C::/48 [200/0]
via 2406:6400::6
B 2406:6400:20::/48 [200/0]
via 2406:6400::7
B 2406:6400:24::/48 [200/0]
via 2406:6400::9
B 2406:6400:28::/48 [200/0]
via 2406:6400::10
B 2406:6400:2C::/48 [200/0]
via 2406:6400::12
Customer p-to-p prefixes in
Route Filtering
Data Centre Prefix
Prefix assignment can be /48 or a number of /48 if need more
/48 is preferred as it will support specific BGP network advertisement
for traffic engineering purpose
Usually advertised in iBGP but ISP can prefer to advertise from
separate AS using eBGP if DC is multihome, has separate routing
policy the IP core and providing public hosting service
Scope within IP core network if single home
For multihoming case origin AS is different and ISP will allow transit
only
Route Filtering
2

B 2406:6400:800::/48 [200/0]
via 2406:6400::2
B 2406:6400:A00::/48 [200/0]
via 2406:6400::5
B 2406:6400:C00::/48 [200/0]
via 2406:6400::8
B 2406:6400:E00::/48 [200/0]
via 2406:6400::11
Data center prefixes in
Receiving Prefixes
There are three scenarios for receiving prefixes from other
ASNs
Customer talking BGP
Peer talking BGP
Upstream/Transit talking BGP
Each has different filtering requirements and need to be
considered separately
Receiving Prefixes: From Customers
ISPs should only accept prefixes which have been
assigned or allocated to their downstream customer
If ISP has assigned address space to its customer, then the
customer IS entitled to announce it back to his ISP
If the ISP has NOT assigned address space to its customer,
then:
Check in the five RIR databases to see if this address space really
has been assigned to the customer. Legitimacy of Address (LoA)
check
The tool: whois h jwhois.apnic.net x.x.x.0/24
(jwhois queries all RIR database)
Receiving Prefixes: From Customers
Example use of whois to check if customer is entitled to
announce address space:
$ whois -h whois.apnic.net 2406:6400::/32

Inet6num: 2406:6400::/32
netname: APNIC-AP
descr: Asia Pacific Network Information Centre
descr: Regional Internet Registry for the Asia-Pacific
descr: 6 Cordelia Street
descr: South Brisbane, QLD 4101
descr: Australia
country: AU
admin-c: AIC1-AP
tech-c: NO4-AP
mnt-by: APNIC-HM
mnt-irt: IRT-APNIC-AP
changed: [email protected]
status: ASSIGNED PORTABLE
changed: [email protected] 20110309
source: APNIC
Portable means its an
assignment to the customer, the
customer can announce it to you
Receiving Prefixes: From Peers
A peer is an ISP with whom you agree to exchange prefixes
you originate into the Internet routing table
Prefixes you accept from a peer are only those they have indicated
they will announce
Prefixes you announce to your peer are only those you have
indicated you will announce
Receiving Prefixes:
From Peers
Agreeing what each will announce to the other:
Exchange of e-mail documentation as part of the peering agreement,
and then ongoing updates
OR
Use of the Internet Routing Registry and configuration tools such as
the IRRToolSet
http://www.isc.org/software/irrtoolset
Receiving Prefixes: From Upstream
Upstream/Transit Provider is an ISP who you pay to give
you transit to the WHOLE Internet
Receiving prefixes from them is not desirable unless really
necessary
Traffic Engineering see BGP Multihoming presentations
Ask upstream/transit provider to either:
originate a default-route
OR
announce one prefix you can use as default

Route Filtering Case study
External Prefixes
Customer prefix coming from downstream
Internet prefix coming from the upstream

Internet
upstream
downstream
AS17821
ISP
Customer
AS65001
Route Filtering
2

B 2406:6400:8000::/48 [20/0]
via FE80::C80C:1FF:FEAF:6, FastEthernet0/0
B 2406:6400:9800::/48 [200/0]
via 2406:6400::3
B 2406:6400:A000::/48 [200/0]
via 2406:6400::4
B 2406:6400:B800::/48 [200/0]
via 2406:6400::6
B 2406:6400:C000::/48 [200/0]
via 2406:6400::7
B 2406:6400:D800::/48 [200/0]
via 2406:6400::9
B 2406:6400:E000::/48 [200/0]
via 2406:6400::10
B 2406:6400:F800::/48 [200/0]
via 2406:6400::12
Downstream customer
prefixes in Training ISP
network:
Route Filtering
2

B 2406:6400:8000::/48 [20/0]
via FE80::C80C:1FF:FEAF:6, FastEthernet0/0
B 2406:6400:9800::/48 [200/0]
via 2406:6400::3
B 2406:6400:A000::/48 [200/0]
via 2406:6400::4
B 2406:6400:B800::/48 [200/0]
via 2406:6400::6
B 2406:6400:C000::/48 [200/0]
via 2406:6400::7
B 2406:6400:D800::/48 [200/0]
via 2406:6400::9
B 2406:6400:E000::/48 [200/0]
via 2406:6400::10
B 2406:6400:F800::/48 [200/0]
via 2406:6400::12
Upstream Internet prefixes in
Route Filtering
Customer prefix coming from downstream:
Option 1: Customer single home and non portable prefix
Customer is not APNIC member prefix received from upstream ISP
Option 2: Customer single home and portable prefix
Customer is APNIC member receive allocation as service provider but no AS
number yet
Option 3: Customer multihome and non portable prefix
Customer is not APNIC member both prefix and ASN received from upstream ISP
Option 4: Customer multihome and portable prefix
Customer is APNIC member both prefix and ASN received from APNIC

Route Filtering
Option 1: Customer single home and non portable prefix

Internet
upstream
downstream
AS17821
Static 2406:6400:8000::/48 to
customer WAN Interface
ISP Prefix
2406:6400::/32

Customer Prefix
2406:6400:8000::/48
NO BGP
Static Default to ISP
WAN Interface
Filter requirement for ISP
Customer interface OSPF passive
No BGP peering with downstream customer
No route filter required
Traffic filter should permit customer prefix only
Filter requirement for Customer
No dynamic routing protocol with ISP
Need traffic filter based on company security policy
Route Filtering
Option 2: : Customer single home and portable prefix

Internet
upstream
downstream
AS17821
Static 2001:0DB8::/32 to
customer WAN Interface
BGP network 2001:0DB8::/32 AS17821 i
Check LoA of Cust prefix

ISP Prefix
2406:6400::/32

Customer Prefix
2001:0DB8::/32
NO BGP
Static Default to ISP
WAN Interface
Static 2001:0DB8::/32 null0

No BGP peering with downstream customer
Traffic filter should permit customer prefix only
No dynamic routing protocol with ISP
Route Filtering
Option 3: Customer multihome and non portable prefix

Internet
upstream
can not change
AS17821
eBGP peering with customer
WAN interface
ISP Prefix
2406:6400::/32

Customer Prefix
2406:6400:8000::/48
BGP peering with downstream customer
Route filter permit 2406:6400:8000::/48 only in
Route filter permit ::/0, AS17821cust, all /48& /32 out
Or route filter permit ::/0 & AS17821 only out
AS path filter permit _64500$ in
Traffic filter should permit customer prefix in
BGP peering with both upstream ISP
Route filter permit 2406:6400:8000::/48 only out
Route filter permit ::/0, AS17821cust, all /48& /32 in
Or route filter permit ::/0 & AS17821 only in
AS path filter permit ^$ out
AS131107
Manual process e-mail to tech-c
Automated process route object or RPKI
Nearly same filter requirement as other ISP
AS64500
eBGP peering with both
ISP WAN Interface
BGP network 2406:6400:8000::/48 AS64500 i
or aggregate address from gateway router
upstream
can change
Route Filtering
Option 4: Customer multihome and portable prefix

Internet
upstream
can change
AS17821
ISP Prefix
2406:6400::/32

Customer Prefix
2001:0DB8::/32
BGP peering with downstream customer
Route filter permit 2001:0DB8::/32 only in
Route filter permit ::/0, AS17821cust, all /48& /32 out
Or route filter permit ::/0 & AS17821 only out
AS path filter permit _64500$ in
Traffic filter should permit customer prefix in
BGP peering with both upstream ISP
Route filter permit 2001:0DB8::/32 only out
Route filter permit ::/0, AS17821cust, all /48& /32 in
Or route filter permit ::/0 & AS17821 only in
AS path filter permit ^$ out
AS131107
Nearly same filter requirement as other ISP
AS64500
eBGP peering with both
ISP WAN Interface
BGP network 2001:0DB8::/32 AS64500 i
or aggregate address from gateway router
upstream
can change
Route Filtering
Downstream Customer BGP In process design issue:
Option 1: ISP default only In
Customer is accepting ::/0 only from upstream ISP prefix
Option 2: ISP default + local In
Customer is accepting ::/0 and upstream ISP prefix and their other customer
portable prefixes (Non portable prefixes should not)
Option 3: ISP default + local + all In
Customer is accepting ::/0, upstream ISP aggregated prefix and their other
customer portable prefixes (Non portable prefixes should not) and all other from
Internet

Route Filtering
Can use a low configuration
router (CPU/DRAM)

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering
router (CPU/DRAM)
Easy to manage small routing
table

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering
router (CPU/DRAM)
table
Do not support destination
specific traffic engineering

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering
router (CPU/DRAM)
table
Can not re-route traffic if remote
transit is down

Internet
Net 2406:6400::/32
upstream
AS17821
default originated
AS131107
default originated
AS64500
::/0 from AS131107
> ::/0 from AS17821
upstream
AS100
AS200
Route Filtering
router (CPU/DRAM)
table
transit is down
I.e. Network 2406:6400::/32 is
withdrawn in AS200 but default
path in AS64500 is still sending
traffic via AS 17821)

Internet
Net 2406:6400::/32
upstream
AS17821
default originated
AS131107
default originated
AS64500
::/0 from AS131107
> ::/0 from AS17821
upstream
AS100
AS200
Route Filtering
router (CPU/DRAM)
table
transit is down
Prefixes originated in AS131107
can be routed via AS17821 (Sub-
optimal path)

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
upstream
AS100
AS200
AS131107
default originated
net 2001:0DB8::/32 i
AS17821
default originated
Route Filtering
Can use a low configuration router
(CPU/DRAM)
Easy to manage small routing table
Do not support destination specific
traffic engineering to the remote
transit is down
AS131107 is sending its portable
local route to AS64500

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from As131107
upstream
AS100
AS200
AS131107
default originated
net 2001:0DB8::/32 i
AS17821
default originated
Route Filtering
Can use a low configuration router
(CPU/DRAM)
Easy to manage small routing table
Do not support destination specific
transit is down
AS131107 is sending its portable
local route to AS64500
can now be routed via AS131107
(Optimal Path)

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from As131107
upstream
AS100
AS200
AS131107
default originated
net 2001:0DB8::/32 i
AS17821
default originated
Route Filtering
Option 3: ISP default + local
+ all In
Need high configuration router
(CPU/DRAM)

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from AS131107
2001:0db8::/32 via AS17821
2406:6400::/32 via AS 131107
> 2406:6400::/32 via AS 17821
etc etc etc
upstream
AS100
AS200
AS131107
default originated
net originated in AS131107
and its portable customer net
AS17821
default originated
Route Filtering
+ all In
(CPU/DRAM)
Need skilled people to manage
large routing table

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from AS131107
2001:0db8::/32 via AS17821
2406:6400::/32 via AS 131107
> 2406:6400::/32 via AS 17821
etc etc etc
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering
+ all In
(CPU/DRAM)
large routing table
Support destination specific

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from AS131107
2001:0db8::/32 via AS17821
2406:6400::/32 via AS 131107
> 2406:6400::/32 via AS 17821
etc etc etc
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering
+ all In
(CPU/DRAM)
large routing table
Can now re-route traffic if remote
transit is down

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from AS131107
2001:0db8::/32 via AS17821
2406:6400::/32 via AS 131107
> 2406:6400::/32 via AS 17821
etc etc etc
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering
+ all In
(CPU/DRAM)
large routing table
transit is down

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from AS131107
2001:0db8::/32 via AS17821
> 2406:6400::/32 via AS 131107
2406:6400::/32 via AS 17821
etc etc etc
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering
+ all In
(CPU/DRAM)
large routing table
transit is down
or AS17821 can now be routed
via AS131107 or AS17821
respectively

Internet
Net 2406:6400::/32
upstream
AS64500
::/0 from AS131107
> ::/0 from AS17821
> 2001:0db8::/32 from AS131107
2001:0db8::/32 via AS17821
2406:6400::/32 via AS 131107
> 2406:6400::/32 via AS 17821
etc etc etc
upstream
AS100
AS200
AS131107
default originated
AS17821
default originated
Route Filtering BCP
Prefixes: From Upstream/Transit Provider
If necessary to receive prefixes from any provider, care is
required.
Dont accept default (unless you need it)
Dont accept your own prefixes
For IPv4:
Dont accept private (RFC1918) and certain special use prefixes:
http://www.rfc-editor.org/rfc/rfc5735.txt
Dont accept prefixes longer than /24 (?)
For IPv6:
Dont accept certain special use prefixes:
http://www.rfc-editor.org/rfc/rfc5156.txt
Dont accept prefixes longer than /48 (?)
Route Filtering BCP
Prefixes: From Upstream/Transit Provider
Check Team Cymrus list of bogons
www.team-cymru.org/Services/Bogons/http.html
For IPv4 also consult:
datatracker.ietf.org/doc/draft-vegoda-no-more-unallocated-slash8s
For IPv6 also consult:
www.space.net/~gert/RIPE/ipv6-filters.html
Bogon Route Server:
www.team-cymru.org/Services/Bogons/routeserver.html
Supplies a BGP feed (IPv4 and/or IPv6) of address blocks which
should not appear in the BGP table
Route Filtering Plan in Training Lab
We will use option 3: Config on ISP Edge router (In)
Receive individual customer prefix
i.e. On R1 From R13 2406:6400:8000::/48
On R3 From R14 2406:6400:9800::/48
On R4 From R15 2406:6400:a000::/48
On R6 From R16 2406:6400:b800::/48
On R7 From R17 2406:6400:c000::/48
On R9 From R18 2406:6400:d800::/48
On R10 From R19 2406:6400:e000::/48
On R11 From R20 2406:6400:f800::/48
And prefix originated by customer AS

We will use option 3: Config on ISP Edge router (Out)
Send default prefix to customer i.e. ::/0
Send aggregated ISP prefix i.e. 2406:6400::/32
Send all individual customer prefix i.e. 2406:6400:8000::/48
2406:6400:9800::/48
2406:6400:a000::/48
2406:6400:b800::/48
2406:6400:c000::/48
2406:6400:d800::/48
2406:6400:e000::/48
2406:6400:f800::/48
Send all Internet prefix with prefix length >/32, /32 and /48 only

We will use option 3: Config on CPE router (IN)
Receive default prefix to customer i.e. ::/0
Receive aggregated ISP prefix i.e. 2406:6400::/32
Receive all individual cust prefix i.e. 2406:6400:8000::/48
2406:6400:9800::/48
2406:6400:a000::/48
2406:6400:b800::/48
2406:6400:c000::/48
2406:6400:d800::/48
2406:6400:e000::/48
2406:6400:f800::/48
Receive all Internet prefix with prefix length >/32, /32 and /48 only

We will use option 3: Config on CPE router (Out)
Send individual customer prefix only
i.e. From R13 To R1 2406:6400:8000::/48
From R14 To R3 2406:6400:9800::/48
From R15 To R4 2406:6400:a000::/48
From R16 To R6 2406:6400:b800::/48
From R17 To R7 2406:6400:c000::/48
From R18 To R9 2406:6400:d800::/48
From R19 To R10 2406:6400:e000::/48
From R20 To R12 2406:6400:f800::/48
Send that prefix originated customer AS number

Questions
Thank you!
End of Workshop

Netsec WS 20131109 BD

Uploaded by

Copyright:

Available Formats

Netsec WS 20131109 BD

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Netsec WS 20131109 BD

Uploaded by

Copyright:

Available Formats

What are some of the main security concepts discussed in the document?

How does the document describe evolving approaches to network security over time?

Network Security

You might also like