Block (data storage)

In <a href="/facts/Computing/Kghtm9fX">computing</a> (specifically <a href="/facts/Data_transmission/1yAhaANd">data transmission</a> and <a href="/facts/Computer_storage/wPHKFpdx">data storage</a>), a block, sometimes called a physical record, is a sequence of <a href="/facts/Byte/BNHUp9Qq">bytes</a> or <a href="/facts/Bit/S972NSaD">bits</a>, usually containing some whole number of <a href="/facts/Record_(computer_science)/1fvSgenB">records</a>, having a maximum length; a block size. Data thus <a href="/facts/Data_structure/yNsNvL1I">structured</a> are said to be blocked. The process of putting data into blocks is called blocking, while deblocking is the process of extracting data from blocks. Blocked data is normally stored in a <a href="/facts/Data_buffer/tVy4j2Dc">data buffer</a>, and read or written a whole block at a time. Blocking reduces the <a href="/facts/Overhead_(computing)/e2jgPrHr">overhead</a> and speeds up the handling of the <a href="/facts/Data_stream/57yDhBCd">data stream</a>. For some devices, such as magnetic tape and <a href="/facts/Count_Key_Data/Gxbr1GwB">CKD disk devices</a>, blocking reduces the amount of external storage required for the data. Blocking is almost universally employed when storing data to <a href="/facts/9-track_tape/5Eb6MjMr">9-track</a> <a href="/facts/Magnetic_tape/wj6IDTR1">magnetic tape</a>, NAND <a href="/facts/Flash_memory/Cn8JsStx">flash memory</a>, and rotating media such as <a href="/facts/Floppy_disk/rShxfuxa">floppy disks</a>, <a href="/facts/Hard_disk/kekLSGSE">hard disks</a>, and <a href="/facts/Optical_disc/vjVDB1GH">optical discs</a>.
Most <a href="/facts/File_system/FdPm90uc">file systems</a> are based on a <a href="/facts/Block_device/t8ISnXft">block device</a>, which is a level of <a href="/facts/Abstraction_(computer_science)/hAnxmJle">abstraction</a> for the hardware responsible for storing and retrieving specified blocks of data, though the block size in file systems may be a multiple of the physical block size. This leads to space inefficiency due to <a href="/facts/Internal_fragmentation/SflHry2p">internal fragmentation</a>, since file lengths are often not integer multiples of block size, and thus the last block of a file may remain partially empty. This will create <a href="/facts/Slack_space/SflHry2p">slack space</a>. Some newer file systems, such as <a href="/facts/Btrfs/lPwmXVJ6">Btrfs</a> and <a href="/facts/FreeBSD/RKsFTm7F">FreeBSD</a> <a href="/facts/Unix_File_System/wDMLrTPe">UFS2</a>, attempt to solve this through techniques called <a href="/facts/Block_suballocation/t2l4N01d">block suballocation and tail merging</a>. Other file systems such as <a href="/facts/ZFS/cmh5BHlp">ZFS</a> support variable block sizes.
Block storage is normally abstracted by a file system or <a href="/facts/Database_management_system/VYcpkevb">database management system</a> (DBMS) for use by applications and end users. The physical or logical volumes accessed via block I/O may be devices internal to a server, directly attached via <a href="/facts/SCSI/pxXulrBr">SCSI</a> or <a href="/facts/Fibre_Channel/C7C0yreb">Fibre Channel</a>, or distant devices accessed via a <a href="/facts/Storage_area_network/e7bmhrmB">storage area network</a> (SAN) using a protocol such as <a href="/facts/ISCSI/oxaQpYBc">iSCSI</a>, or <a href="/facts/ATA_over_Ethernet/GKWzu5S0">AoE</a>. DBMSes often use their own block I/O for improved performance and recoverability as compared to layering the DBMS on top of a file system.
On Linux the default block size for most file systems is 4096 bytes. The stat command part of <a href="/facts/GNU_Core_Utilities/Q9cop5U5">GNU Core Utilities</a> can be used to check the block size.
In <a href="/facts/Rust_(programming_language)/8JNMV19F">Rust</a> a block can be read with the read_exact method.

const BLOCK_SIZE: usize = 4096;

if let Ok(mut file) = File::open("example.bin")
{
    let mut buf = [0u8; BLOCK_SIZE];
    file.read_exact(&mut buf);
}

In <a href="/facts/Python_(programming_language)/YbuGqofa">Python</a> a block can be read with the read method.

BLOCK_SIZE = 4096

with open("example.bin", "rb") as file:
    block = file.read(BLOCK_SIZE)

In <a href="/facts/C_Sharp_(programming_language)/UC2gxeyb">C#</a> a block can be read with the FileStream class.

const int BLOCK_SIZE = 4096;

using FileStream stream = File.Open("example.bin", FileMode.Open);
var block = new byte[BLOCK_SIZE];
await stream.ReadAsync(block, 0, BLOCK_SIZE);

Block (data storage) open-in-new

Block (data storage)