Implementing RAID, LVM, Quota and SAN

File System Management

Whenever we want to manage our file system, we should have the proper information about devices being used whether it is of type “ide”, “scsi”, “sata” or “usb” device. When we are creating a simple “linux” partition or “windows” partition, implementing raid or lvm, using quota, increasing and decreasing the size of “swap” partition, mounting remote file systems, certain things are common. So our first purpose is to properly understand these common things. In this tutorial we have taken great pain in explaining these things in very simple manner.

RAID ((Redundant Array of Inexpensive/Independent Disks)

What is RAID: normally on our systems, we are dividing our hard disk into smaller portions called partitions for various reasons like performance, better management etc. but when we are combining the hard disks, we call the concept RAID. Various types of raids are available, depending upon whether you want speed, fault tolerance or both.

Types of RAID

The type of raid you are going to implement will depends upon the situation and your requirements.

RAID 0
Suppose you have got “900MB” of data. if you are going to save it on 1 drive, it will take time. if we are taking 3 drives and simultaneously trying to save 300MB on each of them. What it will do, it will increase read/write speed. but if 1 drive gets damaged, all data will be lost. That is no fault tolerance. This type of raid is called “RAID 0” ( stripe). So “RAID 0” means speed but no fault tolerance.

RAID 1
Suppose you have got very critical data and you want the data to be there even if your hard disk gets damaged. So we will be taking 2 drives & will implement “RAID 1” (Mirror). Data will get saved on both simultaneously. So even if 1 drive is damaged, still we can access the data. But since we have to write on both disks simultaneously, write speed will be slow. So in this type of raid, fault tolerance is there but less speed.

RAID 5
Suppose we want both speed and fault tolerance, then we have to implement “RAID 5” (stripe with parity). If 1 drive gets damaged, data will still be there. But if another drive fails, then data will be lost. For this type of raid, we need minimum 3 drives.

RAID 6
Suppose we want more fault tolerance than what is available in “RAID 5”, we need to implement “RAID 6” (stripe with dual parity). In “RAID 6”, even if 2 drives are going to fail, data will be there. Minimum 4 drives are required to implement this type of raid.

In this tutorial, we will covering step by step implementation of RAID.

LVM (Logical Volume Management)
What is LVM: Suppose you have created a “/home” partition of size “1200MB”. After some time you realized that allocated space is less, you need more space, ie you want to increase the size of the partition. In another situation, you may be thinking of reducing the size of “/home” partition because very less space is being used and you want to use that extra space for some other purpose. It is in these situations that LVM comes to your help. You can increase/reduce the size of “lvm partitions” according to your needs. It is one of the most important concept in linux.

In this tutorial, we will covering step by step implementation of LVM.

Implementing RAID, LVM, Quota and SAN Tutorial Download

Quota Management
What is quota: Suppose you have got 200GB hard disk on your server and thousands of users are accessing that server & storing their data on the server. Some users might be saving useless stuff on server. After some time you hard disk will run out of space. You want to limit the space used by users.
Suppose you are a web hosting company. You have schemes for individuals/resellers. Since space management is very critical on servers, we want some type of control on being space allocated to users & resellers. Most of the time we are using free mail account services provided by gmail, hotmail, yahoo. When we are accessing these accounts, we find that we have been allocated fixed space. All these things point towards use of quota concept. In linux the quota concept is very advanced. we can implement quota by following criteria
– quota by blocks (size)
– quota by inodes (no. of files)
– quota for users
– quota for groups
– soft/hard limits for users/groups

In this tutorial, we will covering step by step advance implementation of Quota.

Managing Swap Space

What is swap: If system finds that no free RAM is available & it needs more memory, then it would move certain ideal pages to physical hard disk space called “swap” in linux and “virtual memory” in windows, ie swapping enables a system to run larger programs than the available memory (RAM).
When we are installing linux, we are also creating “swap” partition. the size is normally RAM x 2. In our case RAM is 512 so we have given swap partition a size of 1024 MB. now what will happen, if we are going to increase the size of RAM. We need to increase the swap space.

In this tutorial, we will cover swap space management.

LUKS (Linux Unified Key Setup)

LUKS is a disk encryption standard. it is used to encrypt whole block devices. We are covering step by step procedure in this tutorial.

SAN (Storage Area Network) implementation using iscsi

The scsi (small computer system interface) enables computers to communicate with storage devices. The iscsi is used to transfer scsi packets over tcp/ip (ethernet). The terms used & their meanings are
iscsi: internet scsi
iscsi initiator (SAN Client): It functions as iscsi client. It sends the scsi commands over network to iscsi target.
iscsi target (SAN Server): Refers to storage located on iscsi server.
iqn: iscsi qualified name. Used for identifying targets. the format is iqn.yyyy-mm.reverse.domain.name:optionallabel
yyyy: year in 4 digit format (such as 2016)
mm: month in 2 digit format (such as 04)

In this tutorial, we will cover SAN implementation in detail.