My primary concern in learning Linux Red Hat is that I wish to implement
and configure the Linux Web Server in the Internet and Intranet
environment. I've ever read some topics on the Windows NT Web
Server that is called IIS4.0 and was impressed very much at those
wonderful functions they have. Also I want to learn the various
kinds of network components came with Linux by comparing them to the
ones of NT. There are a lot of stuffs I really would like to
master, such as nfs, configuring Linux as a router and proxy server, IP
masquerading, and the amazing stuff of Samba !!! Let's keep
going.....
Written by Gary
Goh........July 2000
** Partitions :
/dev/hdb3 ----the 3rd primary partition on the 2nd IDE hard drive
/dev/sdc6 ----the 2nd logical partition on the 3rd SCSI hard drive
--> The numbers 1,2,3, and 4 denote Primary partitions. The Logical
Partition starts at 5.
To Make Boot(install) disk on DOS or Windows ;
cd\dosutils
d:\dosutils\RAWRITE (enter)
d:\images\boot.img
insert a formatted disk in a drive (enter)
- Each partition is mounted at boot time. The mount process
makes the contents of that partition available as if it were just
another directory on the system. All partitions, when mounted,
appear as a unified directory tree rather than separate drives.
- Linux keeps its list of users in the /etc/passwd
file(,which also contains all the passwords for each user in an
encrypted format). Each system has its own copy of this file
and a user listed in one /etc/passwd file
can not log in to another system unless the user has an entry in the
other /etc/passwd file.
To enable users to log in to any system in a network of computers, Linux
uses the NIS (Network Information System ;
formerly known as Yellow Page) to handle
the remote password file issues.
* Shadow password are a mechanism by
which the actual encrypted password entry is not kept in the /etc/paaswd
file but, rather, in a /etc/shadow file.
The /etc/passwd file remains readable by any user in the system, but
/etc/shadow is readable by the root user only. Good step up in a
security.
* X-Windows is the basis for Linux's graphical user interface.
It is what communicates with the actual video hardware. Programs,
such as KDE and GNOME, use X-Windows as a standard mechanism for
communicating with the hardware.
** Files and Commands
- Using CAT to create and add to Files
CAT stands for "concatenate" meaning to add to the end of,
or to connect and link in a series.
> is known as redirection of standard output.
"Capture the information that normally goes to the screen, create a
file, and put the information into it"
>> is known as appending standard output.
"Capture the information that would normally go to the screen and
append the information to an existing file. If the file
doesn't exist, create it."
< is used to tell the computer "Take the
information from the specified file and feed it to standard input,
acting as though the information is coming from the keyboard."
Use the >> symbol to add data to the end of
a files
cat file1 file2 file3 file4 >
fileout : joins file1, file2, file3, and file4, putting the results in
fileout.
sort files1 >> file2 : the
sort command sorts the contents of files1 and appends (>>) it to a
(perhaps already existing) file2. If file2 doesn't exist, the
system creates file2 and then puts the sorted output into it.
pwd : Print Working Directory
mv afile bfile : This leaves the afile in the
current directory, but changes its name to bfile. (file was not really
moved, but just renamed)
mv afile ../bbdir : This tells Linux to go up one
directory level and look for a directory called bbdir, and then put the
file into that directory with the the name afile. (file really moved)
mv afile ../bbdir/bfile : The afile would have been
moved to the bbdir directory with the name bfile (file moved and
renamed)
mkdir bbdir : This creates a new directory called
bbdir.
rmdir bbdir : To remove the bbdir, the directory
must be empty.
rm -rf bbdir : This removes the bbdir directory and
all files and directories under it.
Relative filenames specify the location of files relative to
where you are.(Ex: ls -la ../../etc/passwd) Filenames that are valid
from anywhere in the file system are called absolute filenames.
(Ex : ls -la /etc/passwd)
** Permissions
1. R (read) for a FILE : read
the file
R for a DIRECTORY : list
the names of the files in the directory ; ls
(R), ls -l (RX)
2. W (write) for a FILE : modify
the file
W for a DIRECTORY : create
or delete files inside the directory
3. X (execute) for a FILE : type and execute
the file ;
(files containing executable Linux commands,
called 'shell scripts' must have both R and X permission by the
person executing them. Programs written in a compiled language
such as C, however, must have only executable permissions, to protect
them from being copied where they shouldn't be copied.)
X for a DIRECTORY : change
the directory with cd command
(Unless you also have R permission for the
directory, ls -l will not work ; ls (R), ls
-l (RX)
ugo rwx (User or Owner, Group, Other Read, Write, eXecute)
Example of changing permissions ; chmod u=r g=rw
o=rwx filename
** File System Management
The command syntax for manually mounting a file system ;
mount block-special-file mount-point
* block-special-file is the device driver
file for the partition of the disk drive (such as a hard drive or a
CD-ROM) where you have made a file system.
* mount-point is the directory where the
file system is mounted.
* An Example : To assign the 1st floppy drive, /dev/fd0, to the /mnt
directory.......
----> mount /dev/fd0 /mnt
The directory where the file system will be mounted (mount-point)
must exist before you issue the mount command.
The command syntax for unmounting
umount name
* the name is either the name of the
block-special-file (/dev/fd0) or the name of the directory that is the
mount-point (/mnt).
Mounting all file systems manually everytime you boot the system
would quickly become annoying. To avoid this, you can have the system
execute the required mount commands when you boot.
The information needed to execute the required mount commands is in the
standard Linux file system configuration file located at /etc/fstab.
The entries in the file usually follow this format ;
block-special-file mount-loc type opts dump-freq
passnumber
EX : /dev/hdb1
/ ext2
default 1
1
After you set up the /etc/fstab configuration
file, file system can be mounted automatically at boot time. And, you
can update the file with a program called linuxconf.
** IP
Masquerade
* Router --> using invalid IP -> IP Masquerading : For
Private Network
--> using valid IP -> No need IP Masq. -> Firewall : For
Security
* Proxy Server
*** The Firewall package is called IP Chains
!! ***
*** Chains is if, then, else structure !! ***
*** IP Masquerading is a form of NAT (Network Address
Translation) !!***
* To set up a Linux computer as a dedicated router with one 3.5"
floppy disk, visit http://www.psychosis.com/linux-router
The Process of Configuring IP Masquerading ;
** IP Masquerading Linux server
(1) Open the netcfg window by typing netcfg&
(2) Click Routing tab
(3) Type ppp0 in the Default Gateway Device and click to select
the Network Packet Forwarding (IPv4)
To verify the above configuration (1)~(3), check and see the network
script in /etc/sysconfig/network, which should be similiar as
below ;
NETWORK = yes
FORWARD_IPv4 = yes
HOSTNAME = rh
DOMAINNAME = win2000.local
GATEWAY =
GATEWAYDEV = ppp0
(4) Run the following commands at shell ;
-> /sbin/ipchains -A forward -j MASQ -s
169.254.0.0/16 -d 0.0.0.0/0
-> /sbin/ipchains -P forward DENY
(5) Run /etc/rc.d/init.d/network restart
You may put the above 2 ipchains commands in /etc/rc.d/rc.local
so that IP Masquerading can be effective at next boot.
** Check Points at IP Masquerading Client computers (Windows9x ,NT, w2k)
;
(1) IP address should be assigned communicable with IP masq.
server computer
(2) Default Gate Way should point at the IP address of IP masq.
server computer.
(3) DNS should be configured correctly, like set it the same IP
of DNS shown in IP masq. server.
** nfs
(Network File System)
** The format of the /etc/exports file is ;
Directory Host(Options)
# Comments
ex : /dosmnt/w98 *.win2000.local #export local file available in
win2000 domain
* The format of Host names in /etc/exports is ;
network address slash sunet or domain
name
ex : 169.254.0.0/255.255.0.0
192.168.18.0/255.255.255.0
*.win2000.local
*pro.win2000.local
???.win2000.local (3 letter
hostnames in win2000.local domain)
* Run the exportfs command after you change the /etc/exports
file
-> /usr/sbin/exportfs -a -v
(-a ; all directories listed in /etc/exports whould
be exported immediately)
(-v ; print verbose output)
* /etc/hosts -------- ip address : host name mapping list
* /etc/host.conf ---- name resolution order
* /etc/resolv.conf --- list of name servers
Stop, Start, Restart programs ---> /etc/ec.d/init.d/nfs
restart
/etc/rc.d/init.d/smb restart (./smb restart)
** RPM (/mnt/cdrom/RedHat/RPMS)
rpm -q samba (query)
rpm -qa
(query all)
rpm -e samba (uninstall)
rpm -ivh samba (install/verify/hash)
** VMWare and Wine
* VMWare ; running guest operating system (Windows) under host
operating system (Linux) ; must have copies of the guest o/s and any
desired application on the hard drive of your Linux system (http://www.vmware.com)
* Wine ; running MS-Dos and Windows application under Linux ;
need not have a copy of MS-Dos or Windows on the hard drive of Linux
system

|