LinuxCBT_EL6_Edition_Notes-part1

###File Types - Permissions###
Features:
 1. Classification of files
 2. Permissions

Tasks:
 1. Classification of files
  a. Use 'ls -l' to expose file properties
'-rw-rw-r--.  1 william william   195127 Jul 18 15:38 vim.jpg'
  '-' -> standard file
'drwxrwxr-x.  4 william william     4096 Jul 23 09:43 cisco'
  'd' -> directory
Note: RHEL6 uses color templates for classifying files:
 'black' -> standard file
 'blue' -> directory
 'red' -> compressed file
 'gree' -> executable
Note: The color pattern is subject to change, so don't always rely upon it
'crw--w----. 1 root root      4,   1 Jul 23 08:54 tty1'
  'c' -> character device
'lrwxrwxrwx. 1 root root          15 Jul 23 08:53 stdin -> /proc/self/fd/0'
  'l' -> symbolic link
'brw-rw----. 1 root disk      8,   0 Jul 23 08:53 sda'
  'b' -> block (storage) device - i.e. hard drive, USB stick, etc.

 2. Permissions
  a. Represented by 9-rightmost bits in 10-bits permissions block
'-rw-rw-r--.  1 william william   195127 Jul 18 15:38 vim.jpg'

'rw-' - owner bits - 2,3,4 = 4+2 = 6
'rw-' - group owner bits - 5,6,7 = 4+2 = 6
'r--' - other/everyone bits - 8,9.10 = 4+0 = 4

Permissions Values:
 'r' = 4 = read
 'w' = 2 = write
 'x' = 1 = execute

  b. Use 'chmod' to influence permissions on file objects - it changes the octal mode
  c. Default permissions are inherited from the $UMASK variable
Note: $UMASK is set in '/etc/login.defs'

  d. 'chmod 666 /tmp/1million.txt'
  e. 'chmod u-w /tmp/1million.txt' - removes owner's ability to write to the content
  e. 'chmod u-w /tmp/1million.txt' - removes other/everyone's ability to write to the content
  f. 'chmod g-w /tmp/1million.txt' - removes group's ability to write to the content
'drwxrwxr-x.  4 william william     4096 Jul 23 09:43 cisco'
Default directory permissions is octal: 775

  g. 'chown linuxcbt /tmp/1million.txt && ls -l /tmp/1million.txt'
  h. 'chmod 0-r /tmp/1million.txt && stat /tmp/1million.txt'
  i. 'chmod 600 /tmp/1million.txt && stat /tmp/1million.txt'
  j. 'chown linuxcbt:linuxcbt /tmp/1million.txt && stat /tmp/1million.txt'
  k. 'chgrp linuxcbt /tmp/1million.txt && stat /tmp/1million.txt'
Note: 'chgrp' 'chown' can change file's ctime (change time)

3. SETUID/SETGID/STICKY Bit
  a. chmod 4755 perl_script_1.pl' - causes script to always run as uer/owner
Note: permission will reflect: '4755' with 'rws'
Note: The 's' replaces the 'x' for the owner to indicate SETUID

  b. 'mkdir /projcetx && chmod 2755 /projectx' - causes files created in directory to inherit group permissions
   b1. 'chmod g+s /projectx'

   b2. 'groupadd projectx'
   b3. 'chown linuxcbt:projectx /projectx'
   b4. 'chmod 2775 /projectx && stat /projectx'

  c. '/tmp' -> example of sticky bit - leading value of: '3'
   c1. 'chmod 3777 /tmp' || 'chmod +t /tmp' - sets sticky bit on object


###Symbolic Links###
Features:
 1. Two types
  a. 'symlinks' - soft - facilitate intra/inter-file-system links
   a1. based on file names in the file system, NOT inode
  b. 'hard links' - hard - facilitate intra-file-system links
   b1. based on inode, NOT file names

Tasks:
 1. Symlinks - shortcuts
  a. 'ln -s /tmp/1million.txt ./tmp.1million.txt'
lrwxrwxrwx.  1 william william       22 Jul 23 11:06 tmp.1million.txt -> gedit-fix-practice.txt
Creates a link to the actual file name
Note: Soft-links do NOT increment the link counter returned by 'ls -l' || 'stat'
Note: So long as the source file name and directory location remain unchanged, the soft-links will work

 2. Hard links - shortcuts to inodes - may not span (go across) file systems
  a. 'ln /tmp/1million.txt /projects/hard.1million.txt' - increments the link counter
  b. 'ls -li /projcets/hard.1million.txt' - reveals inode
Note: permissions apply to ALL linked (hard & soft) file


###Quota Implemenation###
Feature:
 1. Limit storage consumption per user/group
 2. Based on: disk block usage or inode usage
 3. Imposed in 2 stages (thresholds): soft & hard
  a. Soft limit: may be execeeded for up to the grace period
  b. Hard limit: may never be execeeded under any circumstance

Requires:
 1. 'quota*' RPM
 2. Must associates file system(s) with quota management: user and/or group 

Steps:
 1. Enable in: '/etc/fstab'
  a. 'defaults,usrquota,grpquota' - impose on: '/home'
 2. Remount the file system: '/home'
Note: Effect quota management during single-user / installation modes to avoid disconnects in services
  a. 'mount -o remount /home' - remounts the file system
Note: Optional methods of remounting the file system include: umount/mount OR reboot the system
  b. 'mount' - reflects whether or not: 'usrquota', 'grpquota' options have been enabled
 3. Create quota database files and generate disk usage table - defines baseliine
  a. 'quotacheck -cug /home' - applies user and group quotas
Note: 'quotacheck' should be run in: Single-user mode OR when the system reboots to facilitate: read-only remount of target file system
  b. Use: '-m' option to override

 4. Check defined quota database:
  a. 'quotacheck -amvug' - checks quotas - forces check

 5. Assign quota policies per user and/or group:
  a. 'edquota linuxcbt4' - uses default editor($EDITOR)

 6. Run 'quotacheck -avugm' to update stats
 7. Run 'repquota /home' to show FS-wide usage report
 8. Use: 'edquota -t' to modify grace period
  a. 'edquota -T linuxcbt4'

 9 Use: 'quotaon ...' - to enter production mode
  a. 'quotaon -vug /home' - enters production mode
  b. 'quotaon -p /home' - echoes current quota status

Note: Default grace period is 7-days
10. Attempt to write data beyond soft limit grace period


###Provision Partitions & File systems###
Features:
 1. Ability to provision additional storage

Tools:
 1. 'fdisk'
 2. 'parted'
 2. 'make2fs' - ext2, ext3, ext4 FSs

Storage Hierarchy:
 Disk
  -Partition(s)
   -File System(s)

Tasks:
 1. Enumerate available storage:
  a. 'fdisk -l' - enumerates disks and partitions
  b. 'parted -l'

 2. Provision additional storage:
  a. Select disk: /dev/sdb
  b. 'parted /dev/sdb'
  c. 'mkpart primary 1 10GB'
  d. 'mke2fs -t ext4 -j /dev/sdb1' - overlays EXT4 FS on: /dev/sdb1
  e. 'mkdir /temp10G1'
  f. 'mount /dev/sdb1 /temp10G1/ && mount'
  g. Create content in new repository

 3. Repeat process on the same disk

 4. Make partitions available across reboots:
  a. '/etc/fstab'

 4. Unmount both partitions and re-mount via: '/etc/fstab'
  a. 'umount /temp10G1 && umount /temp10G2 && mount'
  b. 'mount -a' - reads the contents of: '/etc/fstab'
Note: Partitioning is typically handled during installation and/or within runlevel 1



###Logical Volume Management (LVM)###
Features:
 1. Volume sets - aggreate storage from disparate sources
 2. Resize storage on-the-fly
 3. Provision storage as necessary


Tasks:
 1. LVM Storage Hierarchy
  Logical Volume - configure file system at this level
   - Volume Groups - represents one or more physicla volumes
    - Physical Volumes: (i.e. /dev/sdb4, /dev/sdc3, etc.) - partition, using fdisk or parted: LVM type (8e)

 2. Create LVM Storage Hierarchy - 6-steps
  a. Create LVM partitions on available disks
   a1. 'parted /dev/sdb' 
   a2. 'mkpart parimary start end'
   a3. 'set partition_num lvm on'
   a4. 'reboot'

  b. 'pvcreate /dev/sdb4 /dev/sdc3' - create physical LVM volumes from partitions
   b1. 'pvdisplay'
  c. 'vgcreate volgroupvar /dev/sdb4 /dev/sdc3' - allocates both volumes to the volume group
  d. 'lvcreate -L 5GB -n logvolvar volgroupvar'
  e. 'mke2fs -t ext4 -j /dev/volgroupvar/logvolvar' - overlays EXT4 FS on LVM volume
  f. 'mkdir /lvmvar1 && mount /dev/volgroupvar/logvolvar /lvmvar1'


###RAID###
Features:
 1. Data spread across 2 or more disk/partitions
 2. Redundancy - recover from catastrophy
 3. Levels: 0,1,4,5,6,10

Tasks:
 1. RAID0
  a. Create multiple partitions: /dev/sd[bc][5-8] - of type '83' || 'linux'
  b. 'init 6' - reboot
  c. 'mdadm --create /dev/md0 --level=0 --raid-devices=2 /dev/sdb5 /dev/sdc5'
  d. 'mke2f -t ext4 -j /dev/md0'
  e. 'mkdir /raid0 && mount /dev/md0 /raid0'
  f. 'nano /etc/fstab'

 2. RAID1 - mirroring - halves the storage
  a. 'mdadm --create /dev/md1 --level=1 --raid-devices=2 /dev/sdb6 /dev/sdc6'
  b. 'mke2fs -t ext4 -j /dev/md1'
  c. 'mkdir /raid1 && mount /dev/md1 /raid1'

 3. RAID5 - striping with parity - sacrifices the equivalent of 1-drive(partition)
  a. 'mdadm --create /dev/md2 --level=5 --raid-devices=4 /dev/sdb7 /dev/sdb8 /dev/sdc7 /dev/sdc8'
  b. 'mke2fs -t ext4 -j /dev/md2'
  c. 'mkdir /raid5 && mount /dev/md2 /raid5 && seq 1000000 > /raid5/1million.txt && ls -l /raid5/1million.txt'
  d. 'nano /etc/fstab'
  e. test auto-mount during system initialization


###RAID Management###
Features:
 1. Create
 2. Assemble: assembles pre-existing array(s)
 3. Manage: use to fail devices to tak them offline
 4. Monitor: E-mail, run processes, etc.
 5. Misc: '--query', '--detail', '--examine'(individual RAID components)

Tasks:
 1. 'cat /proc/mdstat' - enumerates currently-available RAId-arrays (sets)
 2. 'mdadm --query /dev/md[0-2]' - returns information about the 3 arrays: 0-2
 3. Publish RAID array as a read-only volume
  a. 'umount /dev/md0' - unmounts the RAID array
  b. 'mdadm -o /dev/md0' - flags, in the superblock, the array: /dev/md0 as Read-Only
  c. 'mount /dev/md0 /raid0'
  d. 'mount'
 4. Publish RAID array as a read-write volume
  a. 'umount /dev/md0' - unmounts the RAID array
  b. 'mdadm -w /dev/md0' - flags, in the superblock, the array: /dev/md0 as Read-Only
  c. 'mount /dev/md0 /raid0'
  d. 'mount'
 5. Stop RAID volume for management purposes
  a. 'mdadm --manage --stop /dev/md0' - facilitates offline management
Note: Stopping/deactivating the array will remove its '/dev/md?' entry
Note: There are mutiple ways to reassemble RAID arrays:
 1. command-line: 'mdadm -A /dev/md0 /dev/sdb5 /dev/sdc5' - restarts (reassembles) '/dev/md0 from its component parts
 2. '/etc/mdadm.conf' - associates DEVICES & ARRAYS and management/notification info.
  a. 'DEVICE /dev/sdb[5678] /dev/sdc[5678]'
  b. 'ARRAY /dev/md0 devices=/dev/sdb5,/dev/sdc5'

 6. Other options:
  a. 'mdadm -D /dev/md[0-2]' - enumerates info, about ARRAYS
  b. 'mdadm -E /dev/sd[cb][78]' - enumerates info, about the 4 partitions on the 2 drives: /dev/sd[bc]


 3. Resize LVMs
  a. 'lvresize -L 6GB /dev/volgroupvar/logvolvar'
  b. 'resize2fs /dev/volgroupvar/logvolvar 6G'
  c. 'lvresize -L 4GB /dev/volgroupvar/logvolvar'
  d. 'resize2fs /dev/volgroupvar/logvolvar 4G'
Note: Reductions will likely return errors resulting in re-provisioning of the FS


 4. Rename Logival Volume
  a. 'lvrename volgroupvar logvolvar logvolopt' - renames volume, NOT volume group
  b. 'lvresize -L 6GB /dev/volgroupvar/logvolopt' - restores to 6GB

 5. Rename Volume Group
  a. 'vgrename volgroupvar volgroupopt' - renames the volume group
  b. update: /'etc/fstab' - to reflect volume group change

 6. Assign more partitions(storage) to LVM
  a. 'parted /dev/sdc'
  b. 'mkpart primary 16.1GB 26.1GB'
  c. 'set 4 lvm on'
  d. 'pvcreate /dev/sdc4' - assigns LVM partition to LVM management
  e. 'vgextend volgoupopt /dev/sdc4' - extend volume group: 'volgroupopt'
  f. 'lvresize -L 15G /dev/volgroupopt/logvolopt' - online resize
  g. 'resize2fs /dev/volgroupopt/logvolopt 15G' - online resize


 7. LVM GUI
  a. 'system-config-lvm'
  b. 'ssh -X [email protected]' - redirects x.org session back to local GUI
  c. Extend storage of: '/dev/volgroupopt/logvolopt' to: 16GB
Note: GUI will send appropriate commands to system to: 
  a. Resize logical volume (logvolopt)
  b. Resize EXT4 FS to appropriate size

 8. Recoreate LVM hierarchy
  a. Unmount any partitons tied to: '/dev/sd[bc]'
  b. 'parted /dev/sdb' - remove partitiions & create new LVM partitions
  c. 'init 6' - reboot
  d. Use: 'system-config-lvm' to create volume group from: '/dev/sdb1' & '/dev/sdc1'
  e. Create logical volume: 'logvolopt'

###PackageKit###
Features:
 1. GUI for package management
 2. Front-end to YUM
 3. Supports YUM plug-ins

Tasks:
 1. Explore Interface

###Cron###
Features:
 1. Job Scheduler
  a. minutely
  b. hourly
  c. dayly
  d. monthly
  e. yearly
Note: Fields: a-e are specified as per the order above in appropriate config. file

 2. Assumes computer is always on unlike: anacron
 3. Maintains: global and per-user schedules
 4. /var/spool/cron - stores crontabs for: /etc/passwd users or LDAP or otherwise
 5. Checks ALL config files every minute, including: /etc/anacrontab
 6. Supplies 'crontab' utility to manage jobs
 7. Runs in ALL multi-user modes, Does NOT executein: Single-User(1) mode

Tasks:
 1. Analyze current cron setup
  a. 'ps -ef | grep cron'
  b. '/etc/crontab'

 2. Define system-wide job
  a. '*/1 * * * *	linuxcbt	/usr/bin/uptime >> /home/linuxcbt/uptime.stat'

 3. Define per-user job
  b. 'crontab -e' - run as user principle: 'linuxcbt'

 4. Manipulate 'linuxcbt' job as 'root'
  a. 'crontab -e -u linuxcbt' - run as 'root' - edits user's job(s)
  b. 'crontab -l -u linuxcbt' - run as 'root' - lists user's job(s)

 5. Restrict Cron-access
  a. '/etc/cron.allow' - add 'linuxcbt' to list - User Must be on the list in order to submit jobs to 'cron'
  b. '/etc/cron.deny' - add 'linuxcbt2' to list


###Anacron###
Features:
 1. Runs jobs once per day during an allowed interval
 2. Assumes computer is NOT always on, unlike: Cron
 3. Facilitates delays in starting jobs - reduces resource contention
 4. Maintains one schedule: '/etc/anacrontab'

Tasks:
 1. Examine: '/etc/anacrontab'


###Package Management with RPM###
Features:
 1. Compression of packages
 2. SHA-256 hashes are used to sign packages
 3. RPM DB: '/var/lib/rpm' - tracks installed packages, attributes of package file, etc.
 4. 5-Modes of operations:
  a. Install
  b. Uninstall
  c. Upgrade
  d. Query
  e. Verify
 5. Caveat: Does NOT auto-resolve dependencies: Use 'yum'
 6. Caveat: RPM does NOT track non-RPM programs/apps: i.e. '*.tar.gz' || '*.tar.bz2'

Tasks:
 1. Query
  a. 'rpm -qa' - dumps ALL installed packages (RPMs)
  b. 'rpm -qa | grep grep' - 'grep-2.6.3-2.el6.x86_64'
	'grep' - main name of package
	'2.6.3-2' - package version
	'el6.x86_64' - RedHat Version & Platform
  c. 'rpm -qi grep' - returns metadata about 'grep' package
  d. 'rpm -ql grep' - enumerates the contents of the package: 'grep'
  e. 'rpm -qf /bin/grep' - enumerates the file's package membership
  f. 'rpm -qd grep' - enumerates the included documentation
  g. 'rpm -qc lftp' - enumerates a package's configuration file(s)
  h. 'rpm -qpi unix2dos-2.2-35.el6.x86_64.rpm

 3. Verify - Verifies file system contents against installed package in RPM DB
Note: Returns: '.' per test performed if the test passed
Note: If test fails, one of the following will be returned:
5(MD5), S(file size), L(symlink), T(mod time), D(device), m(mode), ?(unreadable file), U(user), G(group)

  a. 'rpm -Vvf /bin/grep' - compares: /bin/grep to 'grep' RPM
  b. 'mv /bin/grep /bin/grep.original && touch /bin/grep'
SM5....T.	/bin/grep

 3. Install - Works if package does NOT exist on the system
  a. 'rpm -ivh unix2dos-2.2-35.el6.x86_64.rpm

 4. Upgrade - Install and/or Upgrades
  a. 'rpm -Uvh unix2dos-2.2-35.el6.x86_64.rpm' 
  b. 'rpm -Uvh --replacepkgs unix2dos-2.2-35.el6.x86_64.rpm

 5. Uninstall
  a. 'rpm -e grep' - checks dependencies and warns where appropriate 

 6. Import RedHat RPM GPG Key to confirm package signatures:
  a. 'rpm --import /etc/pki/rpm-gpg/RPM-GPG-KEY-redhat-release'
  b. 'rpm -qa gpg-pubkey*'