File System In Unix

  • October 2019
  • PDF

This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA


Overview

Download & View File System In Unix as PDF for free.

More details

  • Words: 1,291
  • Pages: 7
1.0

What is a [UNIX System] Administrator?

1. One who administers; one who manages, carries on, or directs the affairs of any establishment or institution. 1.1 Some Common System Administration Tasks. * Performing backups * Adding and removing users * Adding and removing hardware * Restoring files from backups that users have accidentally deleted * Installing new software * Answering users' questions * Monitoring system activity, disc use and log files * Figuring out why a program has stopped working since yesterday, even though the user didn't change anything; honest! * Monitoring system security * Adding new systems to the network * Talking with vendors to determine if a problem is yours or theirs, and getting them to fix it when it is their problem * Figuring out why "the network" (or "Pine", or "the computer") is so slow * Trying to free up disc space * Rebooting the system after a crash (usually happens when you are at home or have to catch a bus) * Writing scripts to automate as many of the above tasks as possible

2.4 Types of Unix users. * root (a.k.a. "super user"--guard this account!) * everybody else 2.5 Becoming root * Log in as root from the system console (avoid) * Use "/bin/su -" command from your regular account (better method) 2.6 Things to be aware of when using root account * $HOME is / (Watch what you delete!) * Change the password frequently and use "good" passwords (more on this later) * Never leave the terminal unattended, even for "just a minute" * Limit who has the root password to as few people as possible * Never execute any regular user's program as root * Never let anyone else run a command as root, even if you are watching them * You risk having your password stolen by "sniffers" if you use programs like "telnet", "ftp", access email remotely. creating Files There are many ways of creating a file. Type cat > Mary_Jones.letter and then type out a few lines of text. You will use this file in later examples. The cat command is used here to write from the keyboard into a file Mary_Jones.letter. At the end of the last line, press Enter one more time and then press Ctrl-D. Now, if you type ls again, you will see the file Mary_Jones.letter listed with any other files. Type cat Mary_Jones.letter without the >. You will see that the command cat writes the contents of a

file to the screen, allowing you to view your letter. It should match exactly what you typed in.

1.1 File systems •

Disk drives are usually read in blocks of 1024 bytes (two sectors). From the point of view of anyone accessing the device, blocks are stored consecutively--there is no need to think about cylinders or heads--so that any program can read the disk as though it were a linear tape.



Now a complex directory structure with many files of arbitrary size needs to be stored in this contiguous partition. This poses the problem of what to do with a file that gets deleted and leaves a data ``hole'' in the partition, or a file that has to be split into parts because there is no single contiguous space big enough to hold it. Files also have to be indexed in such a way that they can be found quickly (consider that there can easily be 10,000 files on a system). UNIX's symbolic/hard links and devices files also have to be stored.



To cope with this complexity, operating systems have a format for storing files called the file system ( fs). Like MS-DOS with its FAT file system or Windows with its FAT32 file system, LINUX has a file system called the 2nd extended file system, or ext2.

1.2 creating a file system 1.2.1 /etc/mkfs /dev/fd0h1440 2400:600 2400 stands for the numbers of 512KB blocks that may be present on the disk. For a 1.44 MB disk the number of blocks can be calculated as,

tracks/side * Number of sides * sectors/track * bytes/sector/512 = 80*2*15*512/512 =2400 600 following the colon indicates the number of inodes that we want to create in the file system. Usually this number is one fourth of the number of blocks.

1.2.2 mke2fs To create a file system on a blank partition, use the command mkfs (or one of its variants). To create a LINUX ext2 file system on the first partition of the primary master run: mkfs -t ext2 -c /dev/hda1 or, alternatively mke2fs -c /dev/hda1 The -c option means to check for bad blocks by reading through the entire disk first. 1.3

Formatting floppies and removable drives

fdformat /dev/fd0h1440

// LINUX

format /dev/rfd0135ds18

//SCO UNIX

rfd stands for raw floppy disk 0 for floppy in drive a 135 tracks per inch ds double sided 18 sector per track fdformat /dev/fd0H1440

// LINUX

1.0

File System Repair: fsck

fsck stands for file system check. fsck scans the file system, reporting and fixing errors. Errors would normally occur only if the kernel halted before the file system was umounted. In this case, it may have been in the middle of a write operation which left the file system in an incoherent state. This usually happens because of a power failure. The file system is then said to be unclean. fsck is used as follows: fsck [-V] [-a] [-t ] <device> -V means to produce verbose output. -a means to check the file system noninteractively--meaning to not ask the user before trying to make any repairs. Here is what you would normally do with LINUX if you don't know a whole lot about the ext2 file system: fsck -a -t ext2 /dev/hda1 although you can omit the -t option because LINUX autodetects the file system. Note that you should not run fsck on a mounted file system. In exceptional circumstances it is permissible to run fsck on a file system that has been mounted read-only. 1.2 packing and unpacking using tar 1.2.1 packing # tar -cvf all.tar /root/rajneesh 1.2.2 Restoring files # tar -xvf all.tar 1.2.3 Listing the files in an archive # tar -tf all.tar 1.2.4 comparing the Files of an archive with disk files

# tar -dvf all.tar 1.2.4 adding

Files to

an archive

# tar -rvf all.tar /root/abc 1.2.4 updating

Files in

an archive

# tar -uvf all.tar /root/abc 1.2.4 deleting # tar -delete

Files from

an archive

-f all.tar /root/abc

1.2.4 concatenating tar

archive

# tar -Af all.tar abc.tar 1.2.4 compressing files # tar -zcvf all.tar.gz /root/test/* 1.2.4 to view compressing files # tar -ztvf all.tar.gz 1.2.5 storing files on external devices # tar -cvpf /dev/fd0 /home/root

1.3

Backing up to floppies

You can use tar to back up to any device. Consider periodic backups to an ordinary IDE drive instead of a tape. Here we back up to the secondary slave: tar -cvzf /dev/hdd /bin /boot /dev /etc /home /lib /sbin /usr /var tar can also back up across multiple floppy disks:

tar -cvMf /dev/fd0 /home/simon 1.4

Tape backups

tar traditionally backs up onto tape drives. The commands

mt -f /dev/st0 rewind tar -cvf /dev/st0 /home rewind scsi tape 0 and archive the /home directory onto it. You should not try to use compression with tape drives because they are error prone, and a single error could make the entire archive unrecoverable. The mt command stands for magnetic tape and controls generic SCSI tape devices. See also mt(1). 1.5 cpio command (i) copy out mode adding the files in the current directory find . -print | cpio -oF /root/abc F option to specify the name of the file (ii) copy in mode cpio -i < /root/abc (iii) copy pass mode -p copy pass mode

to abc

Related Documents

File System In Unix
October 2019 36
File System
October 2019 39
File System
April 2020 32