TRƯỜNG ĐẠI HỌC CÔNG NGHỆ

Bộ môn mạng và truyền thông máy tính

SYSTEM ADMINISTRATION

Lecturer: Dương Lê Minh

[email protected]
FOUNDATION ELEMENTS
 Content
1. Overview
2. Workstations
3. Servers
4. Services
5. Disaster Recovery
6. Security Policy
 Content
1. Overview
2. Workstations
3. Servers
4. Services
5. Disaster Recovery
6. Security Policy
 Overview
• Use a trouble-ticket system
• Manage quick requests right
• Adopt three time saving policies
• Start every new host in a known state
• Our other tips
 Content
1. Overview
2. Workstations
3. Servers
4. Services
5. Disaster Recovery
6. Security Policy
 Definition
• A workstation is a computer hardware
dedicated to a single customer’s work
– Deployed in large quantities
– Have long life cycles (birth, use, death)

………
Life Cycle
 Big Three
• Maintaining workstation operating systems
– Loading the system software and applications
initially
– Updating the system software and
These three tasks
applications should be automated for
any platform that is
– Configuring network parameters
widely used at your site
 Loading the OS
• Automated OS loading
– Jumpstart of Solaris
– KickStart of RHEL
– Remote Installation Service of Microsoft
–…
• Completely vs. partially automatic
• Cloning
Updating the System Software and
Applications
• One, Some, Many
– One. First, patch one machine & make sure
the updates work
– Some. Next, try the patch on a few other
machines (automated!)
– Many. Move to larger and larger groups of
risk-averse customers
 Network Configuration
• DHCP
– Dynamic assignment
– Static assignment
– Lease time assignment
• Automatic scripts
 Content
1. Overview
2. Workstations
3. Servers
4. Services
5. Disaster Recovery
6. Security Policy
 Definition
Workstation Server
• No special hardware • Need special hardware

• Single customer • Multiple customers

• Portable • Reliability & uptime

• 3 years! • Last longer

• Smaller cost • Higher cost

 Server Hardware
Dedicated server Desktop + server
hardware software
Cost High Low
Extensibility V
More CPU performance V
High-performance I/O V
Upgrade options V
Rack mountable V
No side-access needs V
High-availability options V
Maintenance contracts V
Management options V
 Cost of Server Hardware
• 03 product lines
– Home: basic hardware
– Business: up-to-date hardware
– Server: up-to-date
Buy hardware
an + extensibility + …
expensive& Spare
• Maintenance Contract
Server is
• Remote Console Access:
worth!
KVM
 Enhancing Reliability and
Service Ability
• Server Appliances
• Redundant Power Supplies: n + 1
redundancy
• Hot-Swap Components
 Content
1. Overview
2. Workstations
3. Servers
4. Services
5. Disaster Recovery
6. Security Policy
 Definition
• A server is hardware. A service is the
function that the server provides
• Providing a service?
– putting together theending
Never hardware and software
– making the servicestory!
reliable
– scaling the service’s growth
– monitoring, maintaining, and supporting it
 Customer requirements
• The service is being built for the
customers
• Gathering customer requirements
– how to use the new service
– the features they need
– how critical the service will be to them
– what levels of availability and support they will
need for the service
• Consultative process
 Operational Requirements
• SAs: new-service requirements that are
not immediately visible to the customers
• How the service scales
 Open Architecture
• Service from IETF
(RFCs)
– Interoperability
• Service from common
vendors
• The Protocol versus
the Product
 Simplicity
• The simplest solution that satisfies all the
requirements
– most reliable
– easiest to maintain
– easiest to expand
– easiest to integrate with other systems
• Undue complexity
– confusion, mistakes, and potential difficulty of use
– make everything slower
– expensive in set-up cost and maintenance costs
 Vendor Relations
• Use the same hardware and software from
the same reliable vendors
– Updatable & Upgradable (same configuration)
– Price promotion
– Support
 Machine Independence
• The machine should never have a
function-based primary name
When building a
service on a machine,
think about how you
will move it to another
machine in the future
 Environment
• A reliable service needs a reliable
environment
– Availability!
reliable as the
– Data center
weakest link in
– Redundancy the chain of
components
 Restricted Access
• Restrict direct login access on the servers
that are part of the service
– Only SAs for console or remote access
• The more people who log in directly to a
machine, the more likely it is to crash
 Centralization and Standards
• Centralizing the tools, applications, and
services that your customers need
– managed primarily by one central group of
SAs
• Central helpdesk support

 easier to support and lowers training

costs
 Performance & Monitoring
• Customer’s perspective: Does it work, and
is it fast?
• Need performance test
– Tools
• Monitoring for availability, problems, and
performance
– Alarm system
– Log system
 The Icing
• Dedicated machines
– easier to maintain and support
• Full Redundancy
 Content
1. Overview
2. Workstations
3. Servers
4. Services
5. Disaster Recovery
6. Security Policy
 Definition
• A disaster is a catastrophic event that
causes a massive outage affecting an
entire building or site
 Risk Analysis
• Using external consultants
• approximate budget for risk mitigation
(Probable cost of disaster − Probable cost
after mitigation) ×Risk of disaster
 Legal Obligations
• Legal obligations to their
– vendors, customers, and shareholders
in terms of meeting contract obligations
• In DR plan
– how quickly various pieces of the physical and
electronic infrastructure must be restored to
working order
Q&A