Title: CISCO ROUTER AUDIT
1- CISCO ROUTER AUDIT
- Courtesy of and with permission of Ted Schwartz,
Jefferson Wells International
2Email/Sales
7 Application 6 Presentation 5
Session 4 Transport 3 Network 2 Data
Link 1 Physical
7 Application 6 Presentation 5
Session 4 Transport 3 Network 2 Data
Link 1 Physical
Layered Architecture
Topology Architecture
Exec. LAN
Sales Dept.
HR Dept.
IT Dept.
Accounting Dept.
Where and What do POLICIES refer to?
3What is tested in a Audit
Email, DNS, LOGIN, Directory Services, Routing
Table Sharing, SNMP, TFTP, BootP, DHCP,
Applications
Web Servers (internal and External, Accounting
(GL, AP, AR, PR), Human Resources, Groupware,
Transport
Port Scan Session Controls SYN Flood
SSL
Internetwork/Net.
Address Scanning Ping of Death, Ping Flood IP
Address Spoofing
Network Interface/DL
NIC/MAC address spoofing Sniffing
4Enterprise Physical Topological Architecture
Routers C-E-D
Routers F-B-A
5Firewall Architecture
Email Server
External Router
DMZ Zone
Hub
Internal Router
Hub
Internet
Hub
Hub
Intranet
Firewall
6Firewall Architecture
Packet Filtering Router
External Client
Internet
DMZ Zone
Hub
Hub
Firewall
Intranet
7Material Needs
- Obtain these if available Company Network
Policies and printout of router rulebases,
Network Map, List of Network Supported business
applications and network support applications?
Copies of a sample of network logs. A list of
Network security applications virus checker,
firewalls, routers, radius server, TACACS or
RADIUS server, TFTP, SNMP, Active Directory,
Netware Directory Services, Intrusion Detection,
VLANs, VPN, - What business applications are supported by the
network versus being on stand alone servers? Are
they distributed or stand alone? If applications
are distributed are there over-lay maps and
operation descriptions telling distributed
updates - What applications that are listed in the audit
test page does this company use? If used are the
distributed or stand alone applications. Who is
responsible for each application?
8Audit Program Preparation
- Security Policy definition of access allowed to
corporate assets by users and other applications. - Map the users, applications, user of
applications, mangers of applications. - Obtain all distributed overlay network maps with
operations descriptions. If not available and if
this is a full security audit draw and describe
each applications operations than answer these
question on a per application bases. - a. How often is data distributed?
- b. How are updates secured?
- c. Are updates done via VPN?
-
9Audit Notes
- Remember each network conversation is two ways
through a router. - Security Policy definition of access allowed to
corporate assets by users and other applications. - Risk The possible loss or malfunction related
to user of a corporate asset. - Access Control Controlling access to a network
by using network device to limit the type, and
amount of data allowed to be transmitted across
the network. - Intrusion an action taken by someone that is
not allowed access to a network but gets access
for reasons that are not always known. - Detection having a piece of software that
checks the data processing network for actions
taken that are out of the ordinary. This allows
the software to notify management of the
activity. - Multi-Session Applications Applications such as
FTP,HTTP that require multiple sessions to
accomplish their service. - Stay current on network attacks and
vulnerabilities. - Join in security related mailing lists at such
web site aswww.cert.org, www. Securityfocus, and
www.sans.org.
10Access List Section
- How are configurations maintained?
- What are the firewall characteristics used on
each router beyond layer four? - What standby devices exist?
- What is the planning for upgrading the network
capabilities? (VOIP, Video) - What protocols are forwarded that have not been
mentioned? - What applications are supported on the network?
- What protocols are supported by the network and
in what parts are they placed? - What is done towards Virus Management?
- What applications are supported on the network?
- What protocols are supported by the network and
in what parts are they placed? - What are the network security policies and how
are they implemented? - What is done towards Virus Management?
- What is done towards intrusion detection?
11Access List Section
- Are packets denied that have local host,
broadcast, and multicast address. (If any
exceptions please explain. - Are packet s denied that have no IP address?
- Are NFS, Andrew, Xwindows used?
- How are these protocols controlled? ( NTP, SMTP,
DNS, DHCP,SNMP, ICMP, LDAP, BGP, HTTP, LPD,
UUCPD, TFTP, Windows FTP, RPC, POP, IMAP, Netbios
on NT, ICMP, IGMP,RIP, OSPF, EIGRP, ) - What type of access control lists are used?
- What audit procedures are conducted? (Scanning,
log forensics,etc.) - What are the procedures to keep fix and patches
current? - What are current IOS version running?
- Are all of these changes documented?
- Who approves the update process?
- When was the last patch applied?
12Configuration and Change Mgmt
- How is configuration information maintained?
- Are router configurations documented and
authorized by management? - Is the configuration creation method defined and
documented? - Is a history maintained.
- Are vender IOS changes maintained?
- Are fixes and paths implemented?
- What was the last patch implemented?
- Are changes validated or tested?
- Are validations and tests documented?
- Is the processing power of the router enough and
is there enough memory? - Is the a procedure to test and rollout new Cisco
updates?
13Policy Creation
- Are NFS, Andrew, Windows used?
- How are these protocols controlled? ( NTP, SMTP,
DNS, DHCP,SNMP, ICMP, LDAP, BGP, HTTP, LPD,
UUCPD, TFTP, Windows FTP, RPC, POP, IMAP, Netbios
on NT, ICMP, IGMP,RIP, OSPF, EIGRP, JAVA, NAT,
etc.) - Policy process question
- a. Was a site survey done?
- b. How was needed access to external
resources determined? - c. Is a regular review of security
policy needs done? - d. Is a disaster recovery plan in
place that includes the routers? - e. How were router assets identified
and located? - f. How were the standards created for
classifying router policy? - g. How were threat assessment
standards setup? - h Who is responsible for security
policy enforcement at the Cisco router level? - I. How were procedure changes
evaluated related to impact on business and
employees? - Are company security policies keep up to date?
- Are security attack profile kept up to date?
- What are policies related to implementation of
new security technologies? -
14Policy Creation
- Do written policies exist for router use?
- Do the router policies define rules of conduct,
roles and responsibilities? - Do policies define objectives rather than how to
or acls? - Do policies cover multiple levels of security
depending on tasks needing to be accomplished? - Are service and policies that are not stated as
be allow assume to be denied? - Are the network security policies regularly
reviewed? - Is there a security policy defined for physical
damage to the router? - Is the cryptographic algorithm described in a
policy? - Which assets are listed on network policy
documents? - Are software assets identified with users and
user authority? - Do policies spell out the asset, control types,
and authority to change controls? - Who approves the update process? When was the
last patch applied? - Is there stated exactly who can login directly to
the router? - Are standards defined on how to implement
policies? - Do policies define exactly what assets are
protected by the router? - Have policies had a legal review by the legal
department? - Is the person with ultimate authority over router
policy stated in a policy? - Are the network security areas defined in the
remainder of the ICQ spelled out in security
policy?
15Intrusion Detection Audit and Logging
- Are logging methods documented?
- Do alerting and escalation procedures exist?
- Do the procedures exist for 24 hour operation?
- Are advance logging techniques used? (Syslog)
- What is the media that logging archived. (OS)
- Is Cisco IDS implemented on the local routers?
- Are personnel trained in the Intrusion Detection
System? - Does a policy exist for the IDS?
- Are IDS configurations defined for each router?
- Who is authorized to deal with router IDS and
forensics? - Doer support documentation exist for operational
methods logging and forensics? - How are alerts generated for individual
applications review by CBAC? - What and when are audit exceptions investigated?
- How are the exceptions documented?
- What events are audited?
- How long are audit logs kept?
- What tools are used for audit tests?
- Are tools regularly used to test security?
- Is logging configured on exec, commands,
connections and system?
16Intrusion Detection Audit and Logging
- Are router log update sent to a separate
computer? - Is the separate logging computer hardened?
(unnecessary services are disabled) - Is the computer on a trusted network?
- Is logging matched to security policies?
- Is logging reviewed on a regular basis? When was
the later review done for each router? - Are all router configuration changes logged?
- Are all ACL rule results logged?
- Is the time control over logging established and
redundant? - Does your company have a Intrusion Detection
System such a Cisco IDS? - What features does it have? (Alarm and Display
Management, Data Archive, Multiple Level
Management, Centralized Configuration Management,
Notification Modules, and Security Database)
17Password and Access Management
- Are passwords implemented according to
requirements? - Is there a minimum size set for passwords and
what is it? - Are password changes done according to policy?
- Are test done on password strength?
- What is the process set for initial passwords?
- Do users share passwords?
- How are passwords communicated to the user after
being set? - Is a central access authority used? (Radius,
TACACS) - Is TACACS-server notify command used to send a
message when a user makes a TCP connection, logs
out, or enters the enable command. - Is extended TACACS configured?
- How are forgotten passwords dealt with?
- Do router administrator/s understand how to
bypass the enable password? - Is a browser used for router configuration?
- Are routers accessed through remote devices?
(Dialup, Firewall I) - Are exec password put on control and auxiliary
ports? - Has a login banner been created to discourage
inappropriate logins? - Is IPSEC, Kerberos or SSH used for remote
management of the router?
18Physical Security
- Are the routers secured physically?
- Is access to the area restricted to staff that
administrates routers? - Is the physical location locked and alarmed?
- If the router is administrated remotely are those
devices physically secure? - Are alerts issued if entry is made and are the
handled? - Is physical security organized thus preventing
overlooked security weaknesses. - Is a control port used for access
- Is the auxiliary port used for access.
- Is there standby equipment available nearby.
- Are the physical ID numbers listed on a document?
19Specific Protocol Controls
- Is telnet used to administrate the router?
- If telnet is used, make sure access is granted to
only specific nodes. - Is service password encryption
- Is the MD5 encryption used for privileged mode?
- Is CDP disabled on all interfaces?
- Is SNMP used for management?
- If used have community access level password or
community names been changed. - Is SNMP version III used?
- Ensure that Virtual Terminal Timeouts are set.
- If ICMP is used are these blocked on the internet
interface echo in both directions, time
exceeded, redirect, and unreachable - Are inbound packets addressed to the router or
127.0.0.1 on the internal interface dropped and
logged? - Is HTTP used to access the router
- If appropriate is HTTP-access command used to
authorize access to certain addresses. - If DNS is used, only allow DNS traffic to a
specific server. - Are DNS responses allowed to leave the screened
subnet?
20General Audit Questions
- If CBAC is used are the inspection rules used to
deal with FTP, TFTP, etc? - Are inspection rules applied to the appropriate
interface? - Is the console line set to time-out if a user
walks away from a logged in terminal? - Is MD5 encryption used instead of Cisco
proprietary encryption? - Is RIP and OSPF neighbor authentication used?
- How is the key distributed?
- Is a common key used for any group of routers?
- Are any methods used to increase convergence time
in OSPF and RIP. (Convergence increased time
being a security value) - Is the distribution-list command used to suppress
updates from other routers? (OSPF related to
external systems)
21Command Examples
- COMMAND
EXPLANATION - Service password encryption
sets password encryption - No ip finger
disables finger - No ip source route
not allow source routing - Exec-timeout
time out connection - No CDP run
turns off CDP - Access-list list-number (deny/permit) protocol
source source-wildcard source- qualifiers
destination destination-wildcard
destination-qualifiers log -
- (Qualifiers are items that affect the
previously listed command access-list command
such as the source and destination address shown
earlier)
22TCP Termination and ACLs
- TCP termination is critical to the following
Access Control List functions - implemented on Cisco Router. The first ACL
control type is - TCP Intercept will watch for sessions initiated
without an ACK header in response to the SYN
header. It an Cisco router has TCP Inter. Set, it
watches for ACK to SYN relationship and limits
the number requests without an ACK. (This
prevents SYN flood denial of service attacks to a
server) - It limits the number of unacknowledged session to
1100 by default. If it reaches 1100, removes the
oldest session initiation from its table. - It waits 5 seconds after the Fin to terminate a
session allowing for a reset. - Retransmission Time Outs are normally set at one
second. (2,4,8,16, and 32) Under aggressive mode,
time out is halved to .05 seconds and so on. This
is done per one minute sample period. - This is done when Context Based ACLs are
inactive.
23TCP Termination and ACLs
- TCP termination is critical to the following
Access Control List functions - implemented on Cisco Router. The second ACL
control type are - Based upon a session request from a trusted
network, the router waits for the return packets
with the appropriate information swapped. - Reflexive Access Lists base the access through a
router on a session basis. This ACL type operates
on TCP outbound upper layer session information.
Based upon the Acknowledge or Reset bits being
on, the ACL determines if a packet is the first
packet of a session. - It also checks addition session
information such as port and network address when
establishing a session related temporary access
list that will be removed at the end of the
session.
24Lock and Key ACLs
- Lock and Key access lists allow a temporary
access through the firewall after being
authenticated by a name and password. - A telnet session will initiate temporary access
through a router. - After the temporary access is terminated, regular
standard and static extended ACLs are used. - It does not work with multi-channel applications
such as FTP. - It limits the opportunity time for break-ins by
hackers. -
25Content Based ACLs
- The capability is available in the Cisco Firewall
Set. - It creates temporary entries in the appropriate
interface when a session is initiated from a
trusted network. - It inspects control information on control
channels TCP multi-channel applications.
Multi-channel application such as File Transfer
Protocol and H.323. - It does work with UDP session but must
approximate the session state information unlike
the TCP state information that is in a
Transmission Control Block. - Temporary session mean limits on open access and
removal of ACL entries at the end of a session. -