IT3010 Lecture 03: Network Servers

• Network Design and Management is course IT3010
• Lecture 03 focuses on Network Servers

Networked Applications and Architecture

• Networked applications read and write data over a network
• A dominant model to transmit data is through bidirectional, reliable byte streams
• A computer reads and writes to the Internet
• A Web Server reads and writes from the Internet
• The server is a piece of software that manages a shareable resource
• The resource typically resides at a network location where the server runs
• Clients refer to interface software at separate stations, which hide the mechanism for accessing the server
• The Client-server model is a standard for developing network applications
• Server is a process offering a service
• Client is a process requesting a service
• Servers/clients run on potentially different machines
• Servers await requests from clients
• The roles of the client and the server processes are asymmetric

Domain Name System (DNS) Basics

• DNS is a hierarchical, distributed naming system for computers, services, or any Internet/private network resource
• A name server is computer hardware/software that provides responses to directory service queries
• Name servers translate identifiers to numeric identification/addressing components
• Naming schemes identify uniquely entities across an entire network
• Naming is associated with addressing mechanisms beyond a unique identifier

Deep Dive into DNS

• The Internet maintains two principal namespaces: domain name hierarchy and Internet Protocol (IP) addresses
• The DNS maintains the domain name hierarchy and translates services
• The DNS message format contains identification, flags, number of questions, number of answer RRs, number of authority RRs, number of additional RRs, questions, answers, authority, and additional info sections
• Identification of a message matches requests/responses
• Flags are 1-bit each to mark query/response, if authoritative, recursive resolution setting, and its support
• The RR format lists class, name, value, type, and ttl
• Classes are Internet (IN) and Chaosnet (CH)
• Each class determines a value associated with type
• For the IN class:
  • Type A links a name (hostname) to an IP address (value)
  • Type CNAME has the name as an alias for some canonical name and value as a canonical name
  • Type NS has the name as a domain and value as the name of the authoritative name server for the domain
  • Type MX has the value as the hostname of the mail server associated with a name
• Different kinds of mappings are possible, like 1-1, multiple domain names map to the same IP, etc
• Single domain names can map to multiple IPs
• Some valid domain names don't map to any IP addr

DNS Design: Hierarchy and Zone Definitions

• Each node in the hierarchy stores names ending with the same suffix
• The suffix follows a path up a tree
• A 'zone' is a contiguous section of the namespace

DNS Root Name Servers and Servers/Resolvers

• Zones are created by a node convincing an owner to create/delegate a subzone
• Records are stored in multiple redundant name servers
• The primary/master name server gets updated manually
• Secondary/redundant servers receive updates via zone transfer
• Zone transfer is a bulk transfer of the configuration of a DNS server and uses TCP to ensure reliability
• Root name servers are responsible for the root zone
• 13 root name servers exist
• Local name servers contact root servers when they cannot resolve a name
• Each host has a resolver, essentially a library that applications link to for local name servers
• Name servers are responsible for some zone or acts as the local server
• Name servers do lookups of distant host names for local hosts
• Name servers typically answer queries about local zones

DNS Lookup Methods and Message Reliability

• A 'recursive query' has the server search for more information before responding
• 'Iterative query' means a server responds with partial information
• DNS servers are replicated
• Name service is available if at least one replica is up
• Queries can be load balanced
• UDP enables queries, and exponential backoff when retrying servers
• All queries share the same identifier
• It doesn't matter which server responds

Dynamic Host Configuration Protocol (DHCP)

• DHCP enables standardized network protocol used on Internet Protocol (IP) networks
• DHCP dynamically distributes network configuration parameters, such as IP addresses
• Computers use DHCP to request IP addresses and networking parameters automatically from servers
• DHCP reduces the need for manual configuration
• Devices need IP addresses, subnet masks, addresses for default routers and for name servers
• Without DHCP, manual IP address allocation is necessary
• DCHP issues or leases dynamic IP addresses to clients in a network
• It determines Duration, Computer ID etc
• DHCP server assigns/leases IP addresses to a client
• IP addresses will automatically be renewed when clients log into a network
• These IP addresses are taken from a pool of available addresses
• Leases are only distributed to clients with known MAC addresses for improved security
• Additionally, some IP addresses (like those assigned to servers) can be excluded from dynamic assignment
• DHCP follows the steps DHCP discover, DHCP offer, DHCP request and DHCP acknowledge
• Multiple servers may give offer an address

DHCP Message Types

• DHCPDISCOVER
• DHCPOFFER
• DHCPREQUEST
• DHCPACK
• DHCPNAK
• DHCPDECLINE
• DHCPINFORM
• DHCPRELEASE

Understanding auditing

• Network management should start with an audit
• Document and map the network, evaluate and baseline the physical and data link layer infrastructure
• Evaluate and baseline network traffic and protocols and platforms, operating systems and applications

Network mapping and the OSI model

• Network mapping gets to know a network inside-out with detailed descriptions
• The OSI model assists in network mapping
• the OSI model defines a networking framework for implementing protocols in seven layers
• It passes control between layers, starting at the application layer, to the bottom layer
• Common open source mapping tools are Nagios, OpenNMS, and knetmap

Layers of the OSI Model

• Physical Layer:
• Coordinates the functions required to carry bit streams over the physical medium
• Deals with all the mechanical and electrical specifications of interface/transmission media
• Defines procedures and functions for physical devices and interfaces for transmission to occur
• It provides data rate, synchronization of bits, line configuration, physical topology, and transmission mode
• Map physical Topography (actual layout), and physical devices to the model
• Data Link Layer:
• Transforms the physical layer to a reliable link that appears error free
• It's divided into two sub layers: Media Access Control (MAC) and Logical Link Control (LLC)
• The MAC sublayer controls how a computer on the network gains access to data and can transmit it
• The LLC layer oversees frame synchronization, flow control, and error-checking
• Map Framing and physical addressing to the model
• Network Layer:
• It is responsible for source-to-destination delivery of packets with logical addressing and routing
• It ensures that each packet gets delivered from original points to the final destination
• Transport Layer:
• Is responsible for process-to-process delivery of the entire message
• Ensures the whole message is intact and in-order
• Contains service-point addressing, segmentation and reassembly, connection/flow/error control
• Session Layer:
• Acts as the network dialog controller
• Establishes, maintains, and synchronizes the interaction between systems with dialog control and synchronization
• Presentation Layer:
• Concerned with the syntax and semantics of exchanged data involving translation, encryption, and compression
• Identify the type of encryption/compression used
• Application Layer:
• Enables users/software to access the network, support services like mail, remote access, transfers, DB management
• Features network virtual terminals, file transfers, access and managment, mail services, and directory services

Additional Useful Information and Tools

• Non-network information doesn't directly correspond to networking principles yet vital in day to day maintenance
• Key information: Network purpose statement, Network overview documentation, Physical locations, Vendors, Signatories, etc
• Physical Locations requires maintaining Floor plans, Addresses and managers
• Resources include account management, usernames and passwords
• Signatories are the names of who makes decisions and authorize purchases
• Suppliers/Vendors entail having lists of contractors, vendors, service contracts for equipment
• Important mapping tools are Nagios, OpenNMS, knetmap, SmartDraw ™™, Visio, netViz ™™, Neon LANsurveyor

Baselining

• Baselining is optimizing quality of service, measuring, and rating network performance in real-time situations
• Baselining also requires testing and reporting physical connectivity, normal/peak network utilization, throughput
• It finds speed/accessibility problems and vulnerabilities for network analysis
• Provides companies information to make changes or optimize peak performance
• Baselining will determine normal operation, identify/forecast problems, troubleshoot and predict issues
• It will predict scaling and optimization
• Follow a long term, pre-determined schedule. Use during system activity
• Baselining steps:
• Determine what is owned, needs to be measured and when
• Use a measured baseline for future measurement frequency
• Repeat measurements, use alerts and trends to create a data repository
• Recording data during baseline skew results and can cause occur measurement degradation

Core Concepts for Improving Network Performance

• ROI (Return On Investment) is an accounting formula obtaining an actual /perceived future value of an expense investment
• The SPOF ( Single Point Of Failure) is when failure of one generic component of a system malfunction/entire system breaks
• Hardware, electricity, or software components act as potential SPOFs
• MTBF (Mean Time Between Failures) is the average time a device functions before failing
• Disk drives have 500,000 hours of testing before a failure occurs
• MTTR (Mean Time To Repair) is the average time before a component requires service

Baselining Parameters and Proxy Servers

• AFR - Annualized Failure Rate is a the relation between MTBF and the hours a number of devices operate per year
• Uptime is the amount of time the utility is available to users
• Downtime is the amount of time the utility is unavailable to users
• Availability is the percentage the utility is functional
• Proxy Servers provides privacy and caching
• A proxy servers part of a firewall strategy between local/external networks and increases browser performance
• A proxy translates and removes the user's information Types of proxies include:
• Forwarding proxies for target server connections
• Open proxies, accessible for anyone, to conceal IP addresses while on the web
• Reverse proxies, as as surrogate, handles client requests

Web and Mail Servers

• Proxies implement caching functions
• A 'Web Server' has hardware/software to deliver content through the Internet
• HTTP facilitates the transfer of web pages from server to browser
• Virtual Hosting allows one server to be home to multiple sites
• Common servers are Apache, Windows IIS, niginx, GWS
• A Mail Transfer Agent (MTA) transfers email/electronic messages which uses SMTP, POP & IMAP