NETWORK SECURITY
I.
REFERENCES
& RELATED POLICIES
A. Federal & State Regulations
B. University of California
Policies and UC
San Diego Campus Regulations Applying
to Campus Activities, Organizations, and Students
C. Business and Finance Bulletin IS-3:
Electronic Information Security
D. University
of California Electronic Communications Policy
E. Policy and
Procedure Manual (PPM)
|
160-2
|
Disclosure of Information from Student
Records
|
|
460-5
|
Reporting and Investigating Improper
Governmental Activities, Misuse of University Resources, Fraud, and Other
Financial Irregularities
|
|
480-1A
|
Information Within Word Processors, Personal
Computers
|
|
480-1B
|
Personal Computer Backup
|
|
480-3
|
Responsibilities and Guidelines for Handling
Records Containing Information about Individuals
|
|
510-1
|
Use of University Properties
|
Other PPM references
include those dealing with academic,
student, and staff discipline, and use of University equipment for personal financial gain.
F. Personnel Policies for Staff Members
|
62
|
Corrective Action – Professional and Support
Staff (Systemwide)
|
|
62 HR-S-1
|
Corrective Action – Professional and Support
Staff (UCSD)
|
II. PURPOSE
This document sets
forth a security policy
for the UCSD data communications network that will preserve network
integrity and protect
the information assets of network users.
Other University policies also apply to the operation of the campus network. Relevant policies are mentioned under
References above.
III. AUTHORITY
Jurisdiction of
this policy is under the auspices of Administrative
Computing and Telecommunications
(ACT). Questions concerning this
policy should be referred to computingpolicies@ucsd.edu.
IV. DEFINITIONS
A user is any individual who makes use of
the network.
A username is a network user's
personal identifier. Depending on
the computer system this identifier may be variously
known as the Active Directory username,
network user name, Business Systems username, Kerberos principal, or
account name.
A sponsor is an organization or individual who provides verification of a user's need for network services and their
affiliation to the campus.
The UCSD Backbone
network consists of routers,
switches, and cabling that make up the node-to-node campus
network backbone, not including building distribution equipment.
The UCSD Production
network
is all data networking
at UCSD that permits data to flow over
building distribution networks, the UCSD backbone, or a connection to an outside network.
The UCSD Data Communications Network encompasses both the
Production network and all other data networks that may
exist at UCSD. (e.g. experimental
and research networks)
Networking
equipment is equipment
that provides routing, repeating or switching functions on a wired
or wireless network, including routers, switches,
wireless access points, hubs,
and computers that provide any of these
services for a network segment.
Data equipment is equipment
capable of generating binary
data on a wired or wireless network, including, but not limited
to, desktop computers, servers, IP phones, printers, PDAs, netbooks,
tablets, smart phones,
laptops, and IP-based FAX machines.
A server
room
is a room where six or
more ports are dedicated to
servers providing file storage, email, Web and/or other
computing services. Ports in server rooms
are typically provided by a
switch that is owned and managed
by the department and has been
approved by ACT in advance of purchase.
Core network
services are services
such as Active Directory, radius,
VPN, DHCP and DNS that facilitate network
use by attached data and networking
equipment, and that are offered beyond the local switched segment.
Restricted network services are those from which an
authorized user may make
modifications to data, initiate
connections to other networks, or which have data with sensitive or secure
content.
Unrestricted network
services are
those
that provide read-only access to publicly available network services.
V. POLICY
Information is a
principal asset of UCSD and must
be protected from unauthorized
modification, destruction or disclosure, whether accidental or intentional. The UCSD Data
Communications
Network must therefore be kept secure, as it is essential to the transmission
of information. The level of protection on the Network must
be high enough to ensure that the
most sensitive information traversing it is protected while
still allowing free access
to public information.
The security
of the UCSD Data Communications Network, as
a shared resource,
is the responsibility of all network participants. Primary responsibility for the
security of the production network rests with the IT Infrastructure group
of Administrative Computing and Telecommunications
(ACT). All other managers
of a segment of the network (including managers
of networks not directly
connected to the UCSD
Production network) are also responsible for maintaining the security
of their segment.
VI. PROCEDURES
A. Sponsorship
A network user will have at least one sponsor. All
campus users of the network will have an organization
to act as a sponsor;
in most cases this will be the individual's home department. Students are sponsored
by their college. UCSD sponsors limited guest wireless access
to the network, which may be revoked or limited for
violations of Minimum Network Connection
Standards (Exhibit B). Non-UCSD
users of restricted services must find a sponsor within UCSD. Authorization to use services on a network
device is granted by the operator of the service who may also be a sponsor. Sponsors
have the responsibility of ensuring sponsored
users meet all applicable policies,
including Minimum Network Connection
Standards.
B. User Agreement
All users of restricted
network services are bound
by the University
of California Electronic Communications Policy. Key provisions include:
The user agrees to behave in an ethical manner
and will be responsible for his or her own actions. Under
California State Law any person who maliciously accesses, alters,
deletes,
damages or destroys any computer system, network,
computer program or data is
guilty of a felony.
The user understands
that the network is a shared
resource and will not
intentionally take actions that will
interfere with the operation, integrity or security of the network.
The user will not provide access to third
parties without the approval of a
sponsoring organization.
The user understands
that network traffic and files may be subject to search under court
order. In addition, system administrators
may monitor network traffic or access user files as required to protect the integrity of the computer network.
The user understands
that access to the network
may be temporarily suspended during maintenance and that UCSD will not be liable
for damages due to a failure
of some network service or due to a breach of security.
The user should
understand that misuse of networking resources may result in the loss
of privileges. Additionally, misuse can be prosecuted
under applicable
statutes. The user may be held accountable for
his/her conduct under any applicable University
or campus policies,
procedures, or collective bargaining agreements. Complaints
alleging misuse of network resources will be
directed to those responsible for taking appropriate disciplinary action.
C. Network Infrastructure
Any wired or wireless
networking equipment connected to the UCSD Backbone network
must be approved and/or operated by ACT/IT Infrastructure.
Network services
and networking equipment connected
to the UCSD Production network
must be approved by ACT/IT Infrastructure;
the process of obtaining approval should begin before purchase
of said equipment. Such
equipment and services must conform
to current NGN standards and will be operated by ACT/IT Infrastructure except where special
approval is granted, as in the case of server room
switches (see IV Definitions, server room). All
such equipment must provide routing
and access control technology
to enable separation of connected segments
for security and bandwidth control
purposes that meet current campus standards at the time of installation. Installation of networking equipment to avoid the expense of additional wall jacks is not
likely to be approved, nor is any
network equipment to support end
user connections.
Note that ACT/IT Infrastructure provides network firewalls for most UCSD departments or units. Under special circumstances
when centrally-provided
services are unable to meet these needs, department-installed network firewall devices may be approved under the following rules:
·
Each
such device must be pre-approved and properly registered
with ACT/IT Infrastructure,
including the name of a technical contact
for the device,
·
ACT/IT
Infrastructure will be kept informed
of current traffic-filtering rules in
place on each such device,
·
firewall devices must not be used to
obscure downstream hosts (i.e. no NAT),
·
department-installed firewalls will *not* be managed or maintained by ACT/IT Infrastructure.
All networking equipment
must be registered with ACT/IT
Infrastructure, meet all applicable
Minimum Network Connection
Standards (see Exhibit B), permit login access
by ACT Data Communications network personnel and
be located in areas inaccessible
by the general public, secured behind locked
doors, or monitored by staff.
D. Network Connectivity
Anyone wishing
to attach a new piece
of data equipment to the UCSD Production
network will contact ACT/IT Infrastructure prior
to doing so and follow the appropriate ACT/IT Infrastructure registration procedures.
Any attempt to change connectivity
by the introduction of new protocols or new physical
or logical links will be
subject to review by ACT/IT Infrastructure.
Any piece of data equipment
attached to the UCSD Production network
is bound by this policy and
its owner is subject to the policies
listed in Section I, References
and Related Policies. (See Exhibit B, “UCSD Minimum Network
Connection Standards”.)
Departments should,
wherever possible and with
ACT assistance, logically divide their networks between administrative,
research, instruction, or business
activities. In addition, those networks should
be further subdivided by security profile, such as external-facing servers, internal-facing servers, and clients. This allows
for the best combination of performance and security.
E. Service Provision
Providers of network services (e.g. file/print, VPN, proxy, Web services) will do so in a manner that is consistent with good
facility management; network security is a function
of the network participants.
The Minimum Network Connection Standards (Exhibit B) pertaining to the specific classification of server must
be followed. Providers of restricted
network services will ensure that all applicable policies and guidelines (e.g. Business and Finance Bulletin IS-3) for access security are followed.
Services shall
not interfere with other services
on the network.
Services offering access to non-public University resources (network
bandwidth, restricted-access
materials, etc) will authenticate users
in accordance with the terms of use of the resource.
Organizations
will designate a Technical
Contact (a person or group
of people) to be used in the event of
questions or concerns
about a device. It is the organization's
responsibility to keep its contact
data current with ACT/IT Infrastructure.
F. Monitoring, Scanning and Blocking
ACT/IT Infrastructure monitors
traffic on the network for the purpose of maintaining proper network function.
By extension, traffic generated by users
of computer systems on the network is
also monitored. In addition, ACT/IT
Infrastructure uses automated
tools to proactively scan network-attached
devices for vulnerabilities that, if left unaddressed, could
allow those devices to be compromised
and further damage the
network.
ACT/IT Infrastructure practices
proactive blocking. Network devices that are identified as being vulnerable to or having suffered a security breach will be removed from the network at the discretion of ACT/IT Infrastructure until the problem is resolved. Actions
taken for the purposes of
circumventing such a removal
are not acceptable. Devices
may be blocked even when this will
impact devices served by the device in question; security of the campus network is paramount.
Network services
having vulnerabilities that are known to pose a significant threat to campus
network security may be blocked on the network at the discretion of ACT/IT Infrastructure. Solutions
will be available to allow access to
blocked services by authorized
remote users. Exceptions to service
blocks
for individual machines
may be granted where circumstances
dictate. (See Exhibit
A,
“Network Service Port Blocking”)
Any device or equipment that interferes with the campus wired
or wireless network, including, but not limited to, devices creating radio interference and unauthorized wireless
access points, will be removed from the
network at the discretion of ACT/IT
Infrastructure unless ACT/IT Infrastructure has consented in advance to such interference. ACT will attempt to notify the person/group
responsible but if they
are unable to make contact quickly, the violating device may be removed without notice.
G. Remote
Access
In order
to preserve the safety and integrity of the UCSD Data Communications
Network,
UCSD authorized users are
required to use encrypted connections when
connecting remotely to campus
to access, transmit, view or
modify any sensitive data.
The UCSD Virtual Private
Network (VPN) service, administered and managed by
ACT/IT Infrastructure, is the recommended encrypted solution; in some circumstances, other
protocols
such as https and ssh may be sufficient. Users are
advised that use of the VPN for all remote connections is strongly recommended. System or service administrators should use VPN in conjunction with standard
encrypted connection
protocols such as ssh when
administering services using
higher-level (e.g. "root") access.
In special
cases, divisions with higher-security needs
may request ACT approval to
administer local VPN services.
Such approval is contingent upon
the technical inability of
central services to support
requirements of federal, state or local legislation
or the terms of a specific research
contract.
Machines
remotely connected to the UCSD network to conduct University business are expected to conform to the Minimum
Network Connection Standards (Exhibit
B).
H. Miscellaneous
Information regarding major security vulnerabilities and fixes for them is available from ACT. Users should contact their Systems Administrator and/or the manager of their
local network for security
information and local policy.
I. Violations
University policy prohibits
the use of University property for
illegal purposes and
for purposes not in support of the
mission of the University. In addition to
any possible legal sanctions, violators of this policy may be subject to disciplinary action including but
not limited to suspension or dismissal
as relevant, pursuant to
University policies and collective bargaining agreements.
J. Contacts
Questions concerning
this policy should be referred to computingpolicies@ucsd.edu.
Specific questions
about security issues should be directed to security@ucsd.edu.
Questions about
network connectivity or
function should be addressed
to hostmaster@ucsd.edu.
EXHIBIT B
UCSD MINIMUM NETWORK CONNECTION
STANDARDS
1. IMPLEMENTATION
As of January 1,
2010, all standards are in
effect. We have retained effective
dates for historical purposes.
2. MINIMUM
STANDARDS
To connect a device to the campus data communications
network, you must comply with the following standards
and directives.
2.1 Register All
Devices
Register all devices with Administrative Computing and Telecommunications/IT
Infrastructure via the UCSD Hostmaster
(see Appendix D for information).
ResNet machines must be registered
with ResNet instead. Review
registration information periodically and update
it as needed. The registration should
indicate which standard applies
to the device.
2.1.1 Additional Health Sciences Host Registration (Effective January 1, 2008)
In
addition to registration with ACT, register specified Health
Sciences devices with Medical Center
Information Services. Do not
install Life Sustaining Medical Equipment that is dependent upon network
connectivity. If installed on the network, Life
Sustaining Medical Equipment must comply with either section
seven or eight. All servers must
have a current Risk Assessment on file with UCSD MC Information Services. Update this information annually. All
devices must be running a currently-supported operating system.
2.2 Patch
and Update Software
(Effective January 1, 2005)
Campus networked
devices must run software
for which security patches are made
available in a timely fashion.
Review available patches no later
than three days from availability; apply as appropriate.
If a patch is not applied, or cannot be applied for a specific reason, you must apply for an exception with ACT/IT Infrastructure and
comply with all required mitigation.
2.3 Protect Against
Malicious Software (Effective January 1, 2005)
Malicious software
detection and prevention tools appropriate for the
platform, such as anti-virus
software, rootkit
detectors, and system integrity monitoring software,
must be running and kept
up to date. Where machines
are University-owned, the responsibility
for ensuring protective software is updated ultimately rests with the department
or laboratory. The responsibility for these tools on personally-owned devices rests with the individual owner
2.4 Limit
Services (Effective January 1,
2005)
Do not run any
service that is not necessary for the intended purpose or operation
of the device.
2.5 Configure
Host-based Firewall Software (Effective January 1, 2005)
Run and configure
host-based
firewall software to allow
communication only from necessary clients
and only to required services. The presence of an external
access control mechanism does not
obviate the need for host-based firewalls.
Depending on how the device or data
on it is used, UCSD Network
or Data Security groups may require you to install additional protection.
Note that
ACT/IT Infrastructure provides
network-based firewalls to departments and units.
Groups should contact ACT/IT Infrastructure for information on how to apply
that layer of security.
Existence of network firewalls does
not obviate the need for host-based firewalls.
2.6 Use Complex
Passwords (Effective January 1,
2005)
Campus electronic
communications service
providers must have a suitable
process for authorizing any
use of shared restricted electronic communications services
under their control. The mechanism
for providing access to service
users will be referred to here as an
"account".
All campus electronic communications service user accounts
must have either passwords
or another secure authentication system (e.g. biometrics, Smart Cards).
Where possible, devices must be configured
to enforce at least the minimum password complexity requirements specified at the resource found in Appendix
D.
Modify
all default passwords for network-accessible device
accounts, and ensure they are complex. Do not use
the same passwords for privileged and
non-privileged access. Organizations are strongly encouraged to use multi-factor
authentication with appropriate credential
controls for administrative access
to systems.
2.6.1 Additional Health Sciences Password Standards
(Effective January 1,
2008)
Passwords
for administratively privileged accounts must be at least 14 characters
long, unless long passwords are not supported.
2.7 Do
Not Allow Unencrypted Authentication
(Effective January 1, 2008)
All campus devices must use only encrypted authentication mechanisms. In particular, historically insecure
services such as login
via HTTP, Telnet, FTP, SNMP, POP, and IMAP should be replaced
by their encrypted equivalents. In cases where protocols are used without authentication (e.g. HTTP
for general Web pages,
anonymous FTP), use of legacy
protocols is permitted.
2.8 Do
Not Run Unauthenticated
Email Relays (Effective January 1,
2005)
Campus devices must
not provide an active SMTP service that allows unauthorized
third parties to relay email messages,
i.e., to process an e-mail message where neither
the sender nor the recipient is a local user. IP-based authentication is not adequate to meet this requirement. Open email relays will be
removed from the network as soon as they
are detected and without warning.
2.9 Do
Not Allow Uncontrolled
Access to Proxy Services (Effective January 1, 2005)
Proxy
services are not allowed on the campus network unless
they have been approved by
ACT/IT Infrastructure and unless
their configuration and use have been
reviewed and deemed appropriate by that group.
In
particular, software program
default settings in which
proxy servers are automatically enabled must be
identified by the system administrator
and re-configured to prevent uncontrolled access
to proxy services.
Open proxy services will be removed from the network as soon as they are detected and without
warning.
2.10 Enable
Logging (Effective January 1, 2008)
Log all
authentication successes and failures on all devices.
Retain logs for at least
the default retention period for the operating system in use.
2.11 Employ Physical
Security (Effective January 1, 2008)
Where possible and appropriate, configure devices to
"lock" and require a user
to re-authenticate if left unattended for more than 20
minutes. Efforts must be taken to protect computer hardware
and removable media from theft.
2.12 Protect Embedded Data (Effective January 1, 2008)
When a device is decommissioned
or serviced, clean/destroy any internal hard drives according
to the applicable standard. If disk drives are used in these devices for temporary
storage, encrypt data where
possible. Delete data after it
is no longer needed.
3. STANDARDS FOR SHARED PUBLICLY
ACCESSIBLE COMPUTERS
These guidelines cover any publicly accessible machine
that is shared by many different people who may not know
or trust each other. Workstations shared by a group working together are not considered “publicly accessible”
for the purposes of this policy. All devices in this
category must meet the Minimum
Standards outlined above as well as the standards that follow.
3.1 Prevent Accidental Disclosure of Sensitive Information
(Effective January 1, 2008)
In order to prevent the accidental disclosure of sensitive data or the misuse of credentials, devices must prevent persistent storage of files. Any private information must be cleared from the computer between
uses.
Web browser histories
and caches must be cleared.
3.2 Preserve System
Integrity (Effective January 1, 2008)
Verify
and repair the integrity of
the system on a regular basis.
Do not permit changes to the hard disk that could result in unauthorized installation
or modification of software on the computer. Limit access to system features to only those necessary to support the primary function of the
computer.
3.3 Employ Physical
Security (Effective January 1, 2010)
Physically
secure system data cables (e.g. keyboard, network) and their
connections to prevent the insertion of key-logging
or other monitoring hardware by unauthorized persons.
4. STANDARDS FOR PRINTERS,
NETWORK SCANNERS,
NETWORK FAXES, WEBCAMS AND
OTHER NETWORK APPLIANCES
4.1 Restrict Network Access (Effective January 1,
2009)
Deploy all such
devices in private IP space.
Limit network access to authorized entities
(using device- local or
network firewall means).
4.2 Update Firmware
(Effective January 1, 2008)
Apply
firmware updates promptly
when available.
4.3 Protect Embedded Data (Effective January 1, 2008)
When a device is decommissioned
or serviced, clean/destroy any internal hard drives according
to the applicable standard. If disk drives are used in these devices for temporary
storage, encrypt data where
possible. Delete data after it
is no longer needed.
5. STANDARDS FOR
CLIENTS THAT PARTICIPATE IN SENSITIVE ACTIVITIES
These standards cover
any desktop machine where one or more
of the users participate in business or research activities
that expose them to sensitive data. Health
Sciences clinical workstations are considered
“sensitive clients” for the purposes of this document.
See Appendix A for definitions of
data and activities that are considered sensitive.
5.1 General Requirements (Effective
January 1, 2008)
·
All
such devices must meet
the Minimum Standards outlined
above
·
If
unauthorized access is
reasonably believed to have occurred,
the security incident process
of the UCSD Computer Incident
Response Team (CIRT) will be invoked.
5.2 System Configuration
5.2.1 Configure Host-based Firewall
Software to Enforce Client
Status (Effective January 1,
2008)
Configure the
host-based firewall to block
all incoming connections by
default. Incoming connections through
the host-based firewall can be
permitted when they support IT
management and help desk activities, when they apply to specific administrative machines and
specific services, and/or when they allow remote access
through VPN or local network segment
to primary users of the machine.
The host-based
firewall must be a departmentally
or centrally managed firewall product
that logs to a departmental/central
server. We recommend that intrusion prevention capabilities be part
of the host-based firewall product.
5.2.2 Patch
and Update Software
(Effective January 1, 2008)
Apply
security patches within two
weeks of availability.
5.2.3 Protect Against
Malicious Software (Effective January 1, 2008)
Anti-spyware software, if it is available
for the platform, must be run. Available
anti-virus and anti- spyware logs must be reviewed
on at least a weekly basis. Where machines
are
University-owned, the
responsibility for ensuring
protective software is run and logs are reviewed ultimately rests with
the department or laboratory.
The responsibility for these actions on personally-owned
devices rests with the individual owner. To prevent
exposure to malicious software, scan e-mail
file attachments for viruses
and block risky file types (See Appendix
C.)
Web
filtering must be used to prevent exposure to sites that host malicious software.
5.2.4 Enable
Logging (Effective January 1, 2008)
Enable verbose
logging at the operating system
level. Logs must be able to show user, type of event, date and time with time zone, success or failure, and origin
of event, and must identify system
component, affected data, or resource.
In order to allow for event correlation
between different log sources, synchronize clocks using Network Time Protocol (NTP). Set the time source to ntp.ucsd.edu, an Active Directory domain controller, or another accurate time source.
To prevent tampering, push logs off machine
at least weekly to a central log server
and store them for at least two months.
5.2.5 Scan
For Sensitive Data (Effective
January 1, 2008)
Scan system for
unencrypted sensitive data at
least monthly. Where possible, remove sensitive
data from the system. If it cannot be removed, sensitive data must be encrypted.
5.3 User
Management
5.3.1 Use
Secured Authentication
(Effective January 1,
2009)
Authenticate
to an infrastructure
that supports account
fraud detection, authentication
logging, disaster
recovery
and fault tolerance for
system level authentication.
5.3.2 Restrict
Administrative Account Use (Effective January 1,
2008)
User accounts must
not be administrative users, and
administrative access must
only be used when required.
5.4 Vulnerability Management
5.4.1 Vulnerability Scanning
(Effective January 1, 2008)
ACT/IT
Infrastructure will scan devices on
a regularly scheduled basis. Firewall rules must allow for comprehensive
scanning from ACT/IT Infrastructure scanning
machines. Systems must have no real critical vulnerabilities.
5.4.2 Blocking
(Effective January 1,
2008)
In order to protect the sensitive data on these systems, a designated party
will block devices from using the Internet or intranet on
detection of critical vulnerabilities, unless prior arrangements
have been made to mitigate any risk.
5.5 Additional Health Sciences Standards (Effective January 1, 2008)
Only secured
email servers should be
used
to exchange sensitive
data. Non-UCSD e-mail providers
do not meet this standard
unless
approved by Medical Center
Information Services.
6. STANDARDS FOR
SERVERS
THAT PARTICIPATE IN SENSITIVE ACTIVITIES
These standards cover
any servers that host
applications or support clients
that participate in sensitive activities. See Appendix
A for
definitions of data and activities that are considered sensitive.
6.1 General Requirements (Effective
January 1, 2008)
·
All
such devices must meet
the Minimum Standards outlined
above
·
If unauthorized access
is reasonably believed to have occurred the security incident process
of the UCSD Computer Incident
Response Team (CIRT) will be invoked.
6.2 System Configuration
6.2.1 Configure
Host-based Firewall (Effective January 1, 2008)
Configure host-based firewall
software to allow communication only from necessary clients and
only to required services.
To support management, review, and logging, use
a centrally managed and
centrally logging firewall product. Firewall
rules should be supplemented by network ACLs or network-level
firewall rules.
6.2.2 Protect Against
Malicious Software (Effective January 1, 2008)
Use host-based
intrusion-prevention system (IPS) software that can
log and prevent malicious activity. For machines that deal
with large amounts of sensitive data or installations that consist of many systems dealing with
sensitive data, network intrusion detection must also
be used.
Protect
the system with anti-spyware software if it is available for the platform.
6.2.3 Patch
and Update Software
(Effective January 1, 2008)
Apply
security patches within a week
of availability.
6.2.4 Enable
Logging (Effective January 1, 2009)
Enable logging
for the operating system, web server,
and applications that may be running on the server.
Logs must be able to show user, type of event, date and time with
time zone, success or failure, and origin of event, and must identify system component, affected
data, or resource. Review logs regularly, at least three times a week.
To prevent tampering, archive logs to
a central log server or read-only media and restrict
access
to only those with a true business need. Monitor online archived logs with change detection software. Retain logs for at
least three months.
In order to allow for event correlation
between different log sources, synchronize clocks using Network Time Protocol (NTP). Set the time source to time.ucsd.edu, an Active
Directory domain controller, or another accurate time source.
6.2.5 Limit
Services (Effective January 1,
2009)
Use a single server to support
only services related
to a single purpose, rather than offering many generalized services,
in order to limit the potential for
compromise. For example, departmental e-mail can not be hosted on the same server
as departmental personnel's
Web
pages. Services must
run with the least privilege necessary.
6.2.6 Scan
for Sensitive Data (Effective January 1, 2009)
Scan system for
unencrypted sensitive data at
least monthly. Where possible, remove sensitive
data from the system. If it cannot be removed, sensitive data must be encrypted or protected
using another appropriate authorized mechanism.
6.2.7 Manage Users
and Privileged Accounts (Effective
January 1, 2008)
Change any
privileged password when an employee
who knows said password leaves. Make sure that you can associate
activities
performed with elevated privileges with
an identifiable authentication event and specific
individual.
6.3 VULNERABILITY
MANAGEMENT
6.3.1 Vulnerability Scanning
(Effective January 1, 2008)
ACT/IT
Infrastructure will scan devices on
a regularly scheduled basis. Firewall rules must allow for comprehensive
scanning from ACT/IT Infrastructure's scanning
machines. Systems must have no real
high-level vulnerabilities.
6.3.2 Blocking
(Effective January 1,
2008)
In order to protect the sensitive data on these systems, a designated party
will block devices from using the Internet or intranet on
detection of critical vulnerabilities, unless prior arrangements
have been made to mitigate any risk.
6.4
REQUIREMENTS FOR SPECIFIC SERVICES
6.4.1
Web Server
6.4.1.1
Protect Against Malicious Software
(Effective January 1, 2009)
Test
third-party and custom applications for common
web security issues (see the OWASP top ten at http://www.owasp.org/) and repair.
Monitor third-party applications and Web frameworks for patches
and vulnerabilities. Patch any vulnerability within a week.
6.4.1.2
Preserve System Integrity
(Effective January 1, 2009)
In order to detect
compromise or defacement, use change detection
software to monitor static Web
content and Web server configuration for unauthorized
changes.
6.4.1.3
Use Secured Authentication (Effective January 1,
2008)
Where technically possible, use campus Single
Sign-On services to authenticate,
selecting appropriate authentication mechanisms
for the application.
6.4.1.4
Limit Access (Effective January 1, 2008)
Restrict
Web service
to the smallest audience possible, using both authentication and firewall rules.
6.4.2 File
Server
6.4.2.1
Use Secured Authentication (Effective January 1,
2008)
Use non-trivial
authentication to enforce user and
access
control to the network service
and the files within. Restrict access
to authorized clients
and users.
6.4.2.2
Limit Access (Effective January 1, 2008)
Restrict file service to the smallest
audience possible, using
both authentication and firewall rules.
6.4.2.3
Protect Against Malicious Software
(Effective January 1, 2008)
Scan all
shared files for viruses on at least a weekly basis.
6.4.2.4
Encrypt Data Transfer (Effective
January 1, 2008)
Employ
transport-level encryption when transferring unencrypted
sensitive data.
6.4.3 Mail
Server
6.4.3.1
Protect Against Malicious Software
(Effective January 1, 2008)
Employ
technology such as spam filtering and blacklists to limit malicious e-mail delivery.
Block risky file types (see Appendix C). Scan mail folders for viruses at least weekly to
locate e-mail viruses that escaped the initial
scan. Scan all e-mail file attachments for viruses.
6.4.3.2
Encrypt Mail Transport/Delivery (Effective January
1, 2009)
Employ
transport-level encryption between mail clients and
mail servers. Encrypt mail
delivery whenever possible.
7. STANDARDS FOR CLIENTS
CONNECTED TO, OR PART OF, LIFE-SUSTAINING MEDICAL
EQUIPMENT, TREATMENT DELIVERY
SYSTEMS, SUPERVISORY CONTROL AND DATA
ACQUISITION (SCADA), AND HEAVY MACHINERY CONTROL
SYSTEMS
Devices must meet
the standards for “clients
that participate in sensitive
activities” outlined above as well as the standards
below.
7.1 Enable
Logging (Effective January 1, 2008)
Collect logs
as outlined in section 5.2.4, but
retain them for at least six months.
7.2 Scan
for Vulnerabilities (Effective January 1, 2008)
Devices connected
to the campus network or
to the Internet must be scanned weekly using a credentialed Foundstone
scan. Systems must have no real high-level vulnerabilities.
7.3. Protect Against
Malicious Software (Effective January 1, 2008)
To prevent the unauthorized installation of malicious software, Internet-connected
clients must never run as a user with administrative capabilities.
7.4 Configure
Host-based Firewall Software (Effective January 1, 2008)
Devices connected to the campus
network or to the Internet must
have host-based firewall rules that limit
the
outgoing traffic to only established traffic, and limit administrative access to specified administrative machines. Configure
devices to only establish connections with servers connected
to or part of Life-Sustaining Medical Equipment, SCADA, and heavy
machinery control systems
through a bastion host.
Clients may talk directly to servers connected to or part of Life-Sustaining Medical Equipment, SCADA, and heavy machinery control systems
if both systems are exclusively,
and only ever, connected to a common private network.
7.4.1 Additional Health Sciences Standards
(Effective January 1, 2008)
Devices must have
a designated UCSD Health
Sciences IS contact.
8. STANDARDS FOR SERVERS
CONNECTED TO, OR PART OF, LIFE-SUSTAINING MEDICAL EQUIPMENT, TREATMENT
DELIVERY SYSTEMS, SUPERVISORY CONTROL
AND DATA ACQUISITION (SCADA),
AND HEAVY MACHINERY CONTROL
SYSTEMS
Devices must meet
the “standards for servers
that participate in sensitive
activities” outlined above as well as the standards
below.
8.1 Enable
Logging (Effective January 1, 2008)
Collect logs
as outlined in section 5.2.4, but
retain them for at least six months.
8.2 Restrict Network Connectivity (Effective January 1, 2008)
Do not
connect devices directly to
the campus network or the Internet. Configure such
devices so that they cannot directly communicate with any outside host. Necessary communication can be accomplished
using a bastion host to exchange data.
If connected through a bastion host,
ACT/IT Infrastructure
will scan devices on a regularly
scheduled basis using
a credentialed vulnerability scan. Systems must have no real
high-level vulnerabilities.
8.3 Configure
Host-based Firewall Software (Effective January 1, 2008)
Limit
outgoing traffic using the
host-based firewall to only allow necessary
communication with clients meeting the client specification that have no Internet connectivity, and/or with the bastion
host.
8.3.1 Additional Health Sciences Standards
(Effective January 1, 2008)
Devices must have
a designated UCSD Health
Sciences IS contact.
9. STANDARDS
FOR VIRTUAL MACHINE(VM) HOST
SYSTEMS THAT SUPPORT IMAGES OF HOSTS THAT ARE CONNECTED TO, OR
PART OF, LIFE-SUSTAINING MEDICAL
EQUIPMENT, TREATMENT DELIVERY
SYSTEMS, SUPERVISORY CONTROL
AND DATA ACQUISITION (SCADA), AND HEAVY
MACHINERY CONTROL SYSTEMS
Host systems
must meet the “standards for servers
that participate in sensitive
activities” outlined above. Guest
systems must meet the minimum standards.
9.1 Preserve System
Integrity (Effective January 1, 2008)
Verify
and repair the integrity of
the operating system, applications and VM software at least weekly using change
detection software.
9.2 Limit
Services (Effective January 1,
2008)
Do not
offer any services on the host system other than VM management software. Limit access to VM
management software to only
authorized administrative
machines or approved VPN infrastructure as
necessary.
9.3 Restrict Network
Communications (Effective January 1, 2008)
The VM
virtual network interface and host operating system must enforce
network security rules to limit
inappropriate communication between virtual machines and/or
the host.