Transcript ITS CSS Desktop Support

ITS CSS Desktop
Support
Introduction to networking concepts
Last updated: 9/30/2008 by pxahelp
Internet Protocol
IP is used for communicating data
across a packet-switched network.
 Nodes within a network are assigned
an IP address.
 IP acts as a transport layer for other
network protocols (ICMP, TCP, UDP)

MAC Addresses
A MAC Address is a universally
unique identifier (UUID) attached to a
given Network interface (NIC).
 A MAC Address consists of 6 groups
of 2 hexadecimal digits, ex:



01:23:45:67:89:af
DHCP servers often assign IP
Addresses based on MAC.
IP Addresses
An IP consists of 4 “octets”
 Each octet is a binary byte (8-bits),
commonly represented in decimal





00000000 = 0
10000000 = 128
11111111 = 255
The IP address is split into a network
address and host address using a
“subnet mask”
Domain Names
A domain name like claws.rit.edu
corresponds to a single IP Address.
 RIT’s DNS servers:




129.21.3.17
129.21.4.18
DNS servers are provided by DHCP,
and use a simple “query-response”
protocol to resolve DNS names to IP
Addresses.
From Domain Name to MAC





1. DNS query: claws.rit.edu
2. DNS response: 129.21.13.169
3. ARP query: Who has 129.21.13.169? Tell
129.21.233.188
4. ARP response: 00:19:B9:21:06:DD
This is overly simplified, but the actual
discussion is out of scope for this
presentation.
IP Subnetting
Originally the first octet was used to
denote the network and the remainder
were the node. Obviously this is very
inflexible.
 RIT’s network address is 129.21.0.0
 Subnetting was devised as a way for
an organization which controls its own
DHCP and DNS to further divide its
address space.

IP Subnetting
A subnet mask uses the same format
as an IP address, where each bit
who’s value is 1 is part of the network
and everything else identifies the
computer.
 The RIT subnet mask is:


255.255.255.128
IP Subnetting
RIT itself is a “subnet” of the rest of
the internet, however its subnet mask
is 255.255.0.0.
 The remaining 255.128 in binary is 9
bits of 1’s and 7 bits of 0’s.
 So we say that RIT uses a 9-bit
subnet.

9-bit subnetting





What does this mean?
Using 9 bits gives us 512 possible subnets.
(counting 0 and 255)
This is 256 (0-255) values for the 3rd octet
and 256 values for the 4th octet.
Since only a single bit of the subnet mask
applies to the 4th octet only 2 of those 256
values refer to subnets.
At RIT we refer to this division as “high” and
“low” nets. (ex: 233 high means an IP
address above 129.21.233.128)
9-bit subnetting
An IP address whose node address is
empty (all 0s in binary) is considered
to be the subnet ID
 An IP address whose node address is
full (all 1s in binary) is considered to
be the broadcast address for that
subnet.

Meaning?

To target a network (for registration, for
example) you would use



Low: *.*.*.0
High: *.*.*.128
To send something to every machine on
that net (broadcast ping for switch testing)
you would use


Low: *.*.*.127
High: *.*.*.255
Subnet address allocation

In Low Subnets
•
•
•
Addresses 1 - 100 are available for
permanent registrations
Addresses 101 - 119 are reserved
for roaming workstations and are
dynamically allocated by DHCP
Addresses 120 - 126 are reserved
for networking hardware (Fluke,
routers, firewalls)
Subnet address allocation

In High Subnets
•
•
•
Addresses 129 - 228 are available
for permanent registrations
Addresses 229 - 247 are reserved
for roaming workstations and are
dynamically allocated by DHCP
Addresses 248 - 254 are reserved
for networking hardware (Fluke,
routers, firewalls)
Edge cases




In rare instances you will see an 8 bit
subnet. This is distinguishable from a
standard 9 bit net by its subnet mask.
8-bit subnet mask: 255.255.255.0
Using 8 bits instead of 9 creates fewer
subnets with more available host
addresses.
An 8-bit net does not have a high and low.
Its subnet ID is *.*.*.0, like a “low” subnet.
However the broadcast address for one is
*.*.*.255, like a “high” subnet.
Diagnostic Tools: Windows






ipconfig /all – ipconfig will display configuration information for the
computer. ipconfig will display some information adding the /all will
display more information.
ipconfig /release – this will “release” the IP information so the
computer can obtain new information
ipconfig /renew – this will get the new information from the network
for the computer
ping – Sends packets to a specified node to test the connectivity
between two nodes.
tracert – similar to ping, this will send packets to a node on the
network but it reports back information on each node it passes by
to get the node requested.
nslookup – uses DNS to translate an IP into a hostname or a
hostname into an IP.
Diagnostic Tools: Linux/Mac


dig – An interface to the DNS servers, similar to nslookup, but with
far more flexibility and very clear output. Not always available.
ifconfig – InterFace Config. The UNIX version of ipconfig. Displays
broadcast address, subnet mask, current IP, and MAC on a per NIC
basis. ifconfig eth0 down, ifconfig eth0 up is equivilant to ipconfig
/release and ipconfig /renew





Mac does not use eth0 names, use en0 on Mac.
dhclient – Common Linux command (Gentoo, SuSE, Red Hat) to
renew the IP address without disabling the adapter.
nslookup – The same as the Windows version of this command.
Runs a simple query against a DNS server and displays the results.
ping – Mostly the same as the Windows version of this command.
It is important to know that UNIX ping will not stop after 4 pings
unless told. Ctrl+c will quit a running ping.
traceroute – Behaves the same as Windows tracert