Post By: Hanan Mannan Contact Number: Pak (+92)-321-59-95-634 -------------------------------------------------------
DBMS File Structure
Relative data and information is stored collectively in file formats. A file is sequence of records stored in binary format. A disk drive is formatted into several blocks, which are capable for storing records. File records are mapped onto those disk blocks.
File Organization
The method of mapping file records to disk blocks defines file organization, i.e. how the file records are organized. The following are the types of file organization
[Image: File Organization]
Heap File Organization: When a file is created using Heap File Organization mechanism, the Operating Systems allocates memory area to that file without any further accounting details. File records can be placed anywhere in that memory area. It is the responsibility of software to manage the records. Heap File does not support any ordering, sequencing or indexing on its own.
Sequential File Organization: Every file record contains a data field (attribute) to uniquely identify that record. In sequential file organization mechanism, records are placed in the file in the some sequential order based on the unique key field or search key. Practically, it is not possible to store all the records sequentially in physical form.
Hash File Organization: This mechanism uses a Hash function computation on some field of the records. As we know, that file is a collection of records, which has to be mapped on some block of the disk space allocated to it. This mapping is defined that the hash computation. The output of hash determines the location of disk block where the records may exist.
Clustered File Organization: Clustered file organization is not considered good for large databases. In this mechanism, related records from one or more relations are kept in a same disk block, that is, the ordering of records is not based on primary key or search key. This organization helps to retrieve data easily based on particular join condition. Other than particular join condition, on which data is stored, all queries become more expensive.
File Operations
Operations on database files can be classified into two categories broadly.
Update Operations
Retrieval Operations
Update operations change the data values by insertion, deletion or update. Retrieval operations on the other hand do not alter the data but retrieve them after optional conditional filtering. In both types of operations, selection plays significant role. Other than creation and deletion of a file, there could be several operations, which can be done on files.
Open: A file can be opened in one of two modes, read mode or write mode. In read mode, operating system does not allow anyone to alter data it is solely for reading purpose. Files opened in read mode can be shared among several entities. The other mode is write mode, in which, data modification is allowed. Files opened in write mode can be read also but cannot be shared.
Locate: Every file has a file pointer, which tells the current position where the data is to be read or written. This pointer can be adjusted accordingly. Using find (seek) operation it can be moved forward or backward.
Read: By default, when files are opened in read mode the file pointer points to the beginning of file. There are options where the user can tell the operating system to where the file pointer to be located at the time of file opening. The very next data to the file pointer is read.
Write: User can select to open files in write mode, which enables them to edit the content of file. It can be deletion, insertion or modification. The file pointer can be located at the time of opening or can be dynamically changed if the operating system allowed doing so.
Close: This also is most important operation from operating system point of view. When a request to close a file is generated, the operating system removes all the locks (if in shared mode) and saves the content of data (if altered) to the secondary storage media and release all the buffers and file handlers associated with the file.
The organization of data content inside the file plays a major role here. Seeking or locating the file pointer to the desired record inside file behaves differently if the file has records arranged sequentially or clustered, and so on.
Post By: Hanan Mannan Contact Number: Pak (+92)-321-59-95-634 -------------------------------------------------------
DBMS Storage System
Databases are stored in file formats, which contains records. At physical level, actual data is stored in electromagnetic format on some device capable of storing it for a longer amount of time. These storage devices can be broadly categorized in three types:
[Image: Memory Types]
Primary Storage: The memory storage, which is directly accessible by the CPU, comes under this category. CPU's internal memory (registers), fast memory (cache) and main memory (RAM) are directly accessible to CPU as they all are placed on the motherboard or CPU chipset. This storage is typically very small, ultra fast and volatile. This storage needs continuous power supply in order to maintain its state, i.e. in case of power failure all data are lost.
Secondary Storage: The need to store data for longer amount of time and to retain it even after the power supply is interrupted gave birth to secondary data storage. All memory devices, which are not part of CPU chipset or motherboard comes under this category. Broadly, magnetic disks, all optical disks (DVD, CD etc.), flash drives and magnetic tapes are not directly accessible by the CPU. Hard disk drives, which contain the operating system and generally not removed from the computers are, considered secondary storage and all other are called tertiary storage.
Tertiary Storage: Third level in memory hierarchy is called tertiary storage. This is used to store huge amount of data. Because this storage is external to the computer system, it is the slowest in speed. These storage devices are mostly used to backup the entire system. Optical disk and magnetic tapes are widely used storage devices as tertiary storage.
Memory Hierarchy
A computer system has well-defined hierarchy of memory. CPU has inbuilt registers, which saves data being operated on. Computer system has main memory, which is also directly accessible by CPU. Because the access time of main memory and CPU speed varies a lot, to minimize the loss cache memory is introduced. Cache memory contains most recently used data and data which may be referred by CPU in near future.
The memory with fastest access is the costliest one and is the very reason of hierarchy of memory system. Larger storage offers slow speed but can store huge amount of data compared to CPU registers or Cache memory and these are less expensive.
Magnetic Disks
Hard disk drives are the most common secondary storage devices in present day computer systems. These are called magnetic disks because it uses the concept of magnetization to store information. Hard disks consist of metal disks coated with magnetizable material. These disks are placed vertically a spindle. A read/write head moves in between the disks and is used to magnetize or de-magnetize the spot under it. Magnetized spot can be recognized as 0 (zero) or 1 (one).
Hard disks are formatted in a well-defined order to stored data efficiently. A hard disk plate has many concentric circles on it, called tracks. Every track is further divided into sectors. A sector on a hard disk typically stores 512 bytes of data.
RAID
Exponential growth in technology evolved the concept of larger secondary storage medium. To mitigate the requirement RAID is introduced. RAID stands for Redundant Array of Independent Disks, which is a technology to connect multiple secondary storage devices and make use of them as a single storage media.
RAID consists an array of disk in which multiple disks are connected together to achieve different goals. RAID levels define the use of disk arrays.
RAID 0: In this level a striped array of disks is implemented. The data is broken down into blocks and all blocks are distributed among all disks. Each disk receives a block of data to write/read in parallel. This enhances the speed and performance of storage device. There is no parity and backup in Level 0.
[Image: RAID 0]
RAID 1: This level uses mirroring techniques. When data is sent to RAID controller it sends a copy of data to all disks in array. RAID level 1 is also called mirroring and provides 100% redundancy in case of failure.
[Image: RAID 1]
RAID 2: This level records the Error Correction Code using Hamming distance for its data striped on different disks. Like level 0, each data bit in a word is recorded on a separate disk and ECC codes of the data words are stored on different set disks. Because of its complex structure and high cost, RAID 2 is not commercially available.
[Image: RAID 2]
RAID 3: This level also stripes the data onto multiple disks in array. The parity bit generated for data word is stored on a different disk. This technique makes it to overcome single disk failure and a single disk failure does not impact the throughput.
[Image: RAID 3]
RAID 4: In this level an entire block of data is written onto data disks and then the parity is generated and stored on a different disk. The prime difference between level 3 and 4 is, level 3 uses byte level striping whereas level 4 uses block level striping. Both level 3 and 4 requires at least 3 disks to implement RAID.
[Image: RAID 4]
RAID 5: This level also writes whole data blocks on to different disks but the parity generated for data block stripe is not stored on a different dedicated disk, but is distributed among all the data disks.
[Image: RAID 5]
RAID 6: This level is an extension of level 5. In this level two independent parities are generated and stored in distributed fashion among disks. Two parities provide additional fault tolerance. This level requires at least 4 disk drives to be implemented.
