RU2757659C1 - Magnetic drive with divided areas - Google Patents

Abstract

FIELD: computing technology.SUBSTANCE: magnetic drive with divided areas (hereinafter MDDA) is an evolutionary development of an HDD drive and is intended for long-term non-volatile storage of digital information, e.g., in a computer. The principle of operation of the MDDA consists in separate recording of the special sector area to the SSD-mft solid-state memory and the transaction log area to the SSD-log solid-state memory, added to the printed circuit board of the controller of the MDDA. The data sector area therein exclusively occupies all the surfaces of the ferromagnetic layers of the rotating physical disks of the MDDA.EFFECT: technical result is an increase in the reliability of information storage.3 cl, 3 dwg

Description

The technical field to which the invention relates

The magnetic drive with separated areas refers to computer technology, namely to disk non-volatile digital storage devices (CDS) and is intended for repeated rewriting, storage and reading of an array of information in a digital code, for example, on a server or in a computer.

State of the art

TsZU can be roughly divided:

- if possible, record on:

• memory devices, which are written only by the manufacturer (for example, masked ROM chips, CD-ROM).

• memory devices that can be written to by the user using a separate device (for example, the UV erasable EPROM used in early BIOS chips).

• memory, which is written by the end user in the same device that uses it (for example, most types of memory in modern computers).

- if possible rewriting:

• Write-once, non-rewritable (ROM) (eg CD-ROM, CD-R, masked ROM chips).

• Semi-permanent, reprogrammable memory (EPROM) - memory devices with the possibility of multiple rewriting, complicated by long write times or a limited number of write cycles (for example, CD-RW, EPROM chips).

• Devices with free multiple rewriting (for example, hard magnetic disks HDD, memory chips). There is no clear line between this and the previous class.

- by appointment:

• Random access memory (RAM) - memory in which data is located, on which the processor's operations are directly performed. Random access memory can have several hierarchical levels. Examples: SRAM, DRAM.

• Internal long-term storage devices (eg CMOS memory, magnetic HDD, solid state SSD).

• External media intended for backup storage or transfer of information from one device to another (eg floppy disks, flash memory).

• Storage devices for identification and payments (eg magnetic cards, RFID tags).

- by volatility:

• non-volatile, records in which are not erased when the power supply is removed;

• volatile, records in which are erased when the power supply is removed:

статические, которым для хранения информации достаточно сохранения питающего напряжения;

static, for which it is enough to maintain the supply voltage to store information;

динамические, в которых информация со временем разрушается (деградирует), и, кроме подачи электропитания, необходимо производить ее периодическое восстановление (регенерацию).

dynamic, in which information is destroyed (degrades) over time, and, in addition to supplying power, it is necessary to periodically restore it (regenerate).

- by type of access:

• sequential access (eg magnetic tapes).

• random access (RAM; for example, random access memory).

• with direct access (for example, hard drives HDD).

• with associative access (special devices to improve database performance).

- by geometric design:

• disk (magnetic disks, optical, magneto-optical);

• tape (magnetic tapes, punched tapes);

• drum (magnetic drums);

• card (magnetic cards, punched cards, flash cards, etc.);

• printed circuit boards (DRAM cards, cartridges).

- according to the physical principle:

• with magnetic recording

магнитные сердечники (пластины, стержни, кольца, биаксы)

magnetic cores (plates, rods, rings, biaxes)

магнитные барабаны

magnetic drums

жесткий магнитный диск (HDD)

hard magnetic disk (HDD)

гибкий магнитный диск (Floppy)

floppy disk

магнитные ленты (Streamer)

magnetic tapes (Streamer)

магнитные карты (банковские кредитные карты)

magnetic cards (bank credit cards)

• optical about CD

DVD

DVD

HD-DVD

HD-DVD

Blu-ray Disc

Blu-ray Disc

• magneto-optical:

CD-МО

CD-MO

• using the accumulation of electrostatic charge in dielectrics (capacitor chargers, memorizing cathode-ray tubes);

• using effects in semiconductors (EEPROM, flash memory, SSD)

• sound and ultrasonic (delay lines);

• using superconductivity (cryogenic elements);

- by the number of stable states:

• binary

• ternary

• decimal

The information capacity of binary CPUs is measured in bits, and ternary CPUs are measured in trites.

The main parameters of digital memory include information capacity, power consumption, storage time, and performance.

The most commonly used non-volatile direct-access memory devices include "Hard Disk Drive" and "Solid State Drive", as well as "Hybrid Hard Drive":

--- "Hard disk drive" (hereinafter referred to as HDD-drive), also called a hard disk, HDD, HDD, HMDD, HDD, hard (magnetic) disk drive or in slang "hard drive". Its principle of operation is based on the interaction of the magnetic field of a stationary magnetic head and a thin layer of ferromagnet, which is displaced, like in a tape recorder, relative to the magnetic head.

The design of the HDD is shown in Fig. 1 and includes:

• valve motor spindle 1

• rotating physical disk 2, which is fixed on the valve motor spindle 1 and covered with a thin layer of ferromagnet. The ferromagnet is protected from the mechanical effect of the magnetic head by a layer of carbon and a layer of polymer. The total layer thickness is about 12 nm. Typically, multiple physical hard disks are attached to a spindle.

• rocker arm 3, which rotates on the axis of rocker arm 4 and at the ends of the levers of which are fixed:

звуковая катушка коромысла 5, которой осуществляется дискретное поворачивание коромысла в необходимое положение за счет ее взаимодействия с постоянным магнитом 6

the voice coil of the rocker arm 5, which performs discrete rotation of the rocker arm to the required position due to its interaction with the permanent magnet 6

слайдер 7 - деталь с особой аэродинамической формой, которая закрепляется на рычаге коромысла около магнитной головки 8. Поток воздуха от вращающегося физического диска 2 создает в слайдере подъемную силу, которая поднимает магнитную головку 8 на высоту около 10 Нм относительно слоя ферромагнетика.

slider 7 is a part with a special aerodynamic shape, which is fixed on the rocker arm near the magnetic head 8. The air flow from the rotating physical disk 2 creates a lifting force in the slider, which raises the magnetic head 8 to a height of about 10 Nm relative to the ferromagnetic layer.

магнитная головка 8 записи/чтения данных, например, на основе туннельного магниторезистивного эффекта с оксидом магния. Магнитные головки находятся над каждой стороной всего пакета вращающихся физических дисков 2

magnetic head 8 for writing / reading data, for example, based on the tunnel magnetoresistive effect with magnesium oxide. Magnetic heads are located above each side of the entire stack of rotating physical disks 2

коммутатор-предусилитель 9, который закреплен на коромысле 3 около магнитной головки 8. Создаваемый магнитной головкой 8 сигнал очень слаб и для его повышения должен усиливаться до уровня цифровых сигналов

switch-preamplifier 9, which is fixed on the rocker 3 near the magnetic head 8. The signal created by the magnetic head 8 is very weak and to increase it must be amplified to the level of digital signals

• printed circuit board 10 of Fig. 2, into which microcircuits are soldered, for example:

основной контроллер 11, который обрабатывает инструкции, транслирует потоки данных внутрь и наружу, корректирует ошибки, управляет током звуковой катушки коромысла 5 и т.п., например, LSI В64002

the main controller 11, which processes instructions, translates data streams in and out, corrects errors, controls the current of the voice coil of the rocker arm 5, etc., for example, LSI B64002

КЭШ-память 12 (или буферная память) - это особая разновидность энергозависимой оперативной памяти, своеобразная «прослойка» между магнитным диском и другими компонентами компьютера, в которой временно хранятся и обрабатываются часть данных HDD-накопителя. Предназначена для более плавного считывания информации и хранения данных, к которым на текущий момент чаще всего обращается пользователь или операционная система. Например, микросхема Samsung К4Т51163QJ: 64 МБ DDR2 SDRAM с тактовой частотой 800 МГц.

Cache memory 12 (or buffer memory) is a special type of volatile random access memory, a kind of "layer" between the magnetic disk and other computer components, which temporarily stores and processes part of the data from the HDD. Designed for smoother reading of information and storage of data, which at the moment is most often accessed by the user or the operating system. For example, the Samsung K4T51163QJ chip: 64 MB DDR2 SDRAM with a clock frequency of 800 MHz.

микросхема управления вентильным мотором 14, например Smooth MCKXL

valve motor control chip 14, for example Smooth MCKXL

последовательная флеш-память 15, которая используется для хранения встроенного программного обеспечения HDD-накопителя (немного похожего на BIOS компьютера), например, Winbond 25Q40BWS05 объемом 500 КБ

15 serial flash memory, which is used to store the firmware of an HDD drive (a bit like a computer's BIOS), such as a 500KB Winbond 25Q40BWS05

разъемы для подключения питания 16 и информационного кабеля 17, например, интерфейса SATA.

connectors for connecting power 16 and data cable 17, for example, SATA interface.

Several levels are used to work with disks:

• physical layer of access to data on the disk using the disk controller

• access to data using BIOS (interrupt 13h)

• access to data using the means of operating systems

• access to data using the means of high-level languages.

During the factory low-level formatting of the disk in a ferromagnet, the surface of the disk is formed by the magnetic head 8 concentric tracks (track).

Each track consists of a set of sectors. The standard capacity of one sector is 512 bytes (4096 bits) and cannot be logically or physically split.

Tracks are numbered from the edge to the center of the disc, starting at 0. Sectors are numbered within one track, starting at 1. Sides are numbered starting at 0.

It is accepted to logically combine several sectors into a cluster (cluster). The number of sectors in one cluster depends on the total size of the HDD. It is impossible to write information from different files into one cluster. The cluster is not subject to logical separation.

All tracks located on different physical disks, but at the same distance from the center, are called a cylinder. The cylinder number corresponds to the numbers of the tracks included in this cylinder. As information is recorded, the cylinders gradually fill towards the center of the disc. Within a cylinder, the transition from one track of the current disc to a track of another disc does not require moving the read / write heads. Thus, by optimizing the movement of the heads, a decrease in the time for moving the heads from one sector to another is achieved.

In order to unambiguously indicate the addresses of the data block, it is necessary to define three numbers:

• cylinder number (track is determined)

• number of the magnetic head (the number of the physical disk and its side is determined)

• sector number on the track.

The so-called absolute disc numbering is used here.

However, the operating system considers all sectors of the disk as one contiguous array, the elements of which are numbered in a row, starting from 0, therefore, each sector, except for the absolute numbering, has its own logical or relative number.

Tracks are grouped one by one (for example, drive C) or several partitions (logical drives).

The organization of several partitions provides the ability to install on each of them its own operating infrastructure with a separate volume and logical disks.

All logical drives are provided with a volume boot sector, two copies of the file allocation table, and a root directory. Through the above structures, the operating system manages to allocate disk space, track the location of files, and bypass damaged areas on the disk.

High-level formatting comes down to creating partitions, disk tables of contents and file system (FAT, NTFS, etc.). With the DOS FORMAT command, a floppy disk undergoes both types of formatting at once, while a hard disk only undergoes high-level formatting.

The tracks and sectors of the rotating physical disc 2 of FIG. 1 are broken down into:

• the parking area of the magnetic heads 20, where the magnetic head 8 "lands" before turning off the HDD-drive, which prevents the magnetic heads from damaging the ferromagnet layer of the rotating physical disk 2.

• area of boot sectors 21 with the address "side 0, track 0 and sector 1". In general, this sector contains the address of the start of the special sector area, but may contain a bootstrap program. It also contains a Partition Table, where one of the four partitions is marked as active, and control is transferred to the directory of the boot program. The Partition Table indicates the location of the LUN sector blocks

• the area of special sectors in which the file system is located, for example:

область специальных секторов FAT32 22 фиг. 1, которая применялась в Windows до 1993 года. Поддерживает диски размером до 128 Гб, а размер отдельных файлов - в пределах 4 Гб. Область специальных секторов файловой системы FAT32 располагается в начале вращающегося физического диска, сразу после области загрузочных секторов 21.

FAT32 special sector area 22 of FIG. 1, which was used in Windows until 1993. Supports drives up to 128GB and individual files up to 4GB. The FAT32 file system special sectors area is located at the beginning of the rotating physical disk, immediately after the boot sectors area 21.

область специальных секторов 23 фиг. 1 ОС NTFS - применяется в начиная с серверов NT и Windows Vista. Поддерживает диски размером до 16 Тб.

the area of special sectors 23 of FIG. 1 NTFS OS - used in NT and Windows Vista servers. Supports drives up to 16 TB.

The area of special sectors 23 of the NTFS file system is located on tracks in the middle of a rotating physical disk and contains, for example:

корневой каталог

root directory

MFT (Master File Table). NTFS резервирует первые 16 записей таблицы размером около 1 Мб для специальной информации

MFT (Master File Table). NTFS reserves the first 16 records of the table, about 1 MB in size, for special information

Зеркальную запись MFT. Если первая запись MFT разрушена, NTFS считывает вторую запись

Mirror record MFT. If the first MFT record is corrupted, NTFS reads the second record

Семнадцатая и последующие записи главной файловой таблицы используются собственно файлами и каталогами на томе

The seventeenth and subsequent records of the master file table are used by the actual files and directories on the volume

Файл битовой карты (bitmap file), где хранится схема распределения пространства на томе. Атрибут данных этого файла содержит битовую карту, каждый бит которой представляет один кластер тома и указывает, свободен ли данный кластер или занят некоторым файлом

A bitmap file that stores the allocation of space on a volume. The data attribute of this file contains a bitmap, each bit of which represents one cluster of the volume and indicates whether the given cluster is free or occupied by some file

Файл плохих кластеров (bad cluster file) для регистрации поврежденных участков на томе

Bad cluster file for registering bad sections on a volume

Файл тома (volume file), содержащий имя тома, версию NTFS и бит, который устанавливается при повреждении тома

Volume file containing the volume name, NTFS version and bit that is set when the volume is damaged

Файл (attribute definition table), содержащий таблицу определения атрибутов, которая задает типы атрибутов, поддерживаемые на томе, и указывает можно ли их индексировать, восстанавливать операцией восстановления системы и т.д.

An attribute definition table file that contains an attribute definition table that specifies the types of attributes supported on the volume and whether they can be indexed, restored by a system restore, and so on.

exFAT (FAT64) - поддерживает все актуальные системы, в том числе Мас и Linux. Работает с файлами размером более 4 Гб и рассчитана на использование в съемных устройствах.

exFAT (FAT64) - Supports all current systems, including Mac and Linux. Works with files larger than 4 GB and is designed for use in removable devices.

HFS+ -поддерживает исключительно устройства Apple

HFS + - only supports Apple devices

• data sector area 24 of FIG. 1, which contains the files stored on the HDD-drive information

• transaction log area 25 of FIG. 1, which logs all operations affecting the volume structure to NTFS, including file creation and any commands that change the directory structure. The Transaction Log area records the "Transaction Log", which is used to recover an NTFS volume after a system failure. The transaction log is a regular, non-fixed length log file that is written to an arbitrary location in the data sector area 24 of FIG. 1.

In the clusters of the data sector area 24 of FIG. 1 files are written according to the file system MFT table, for example:

• Driver files

• Operating system files

• Paging file, which is used when there is insufficient RAM

• Application files

By dividing information into files, it is possible for a machine and a person to distinguish one set of data from another when searching for, changing, deleting or performing other operations with them.

It should be noted that it is important to maintain the integrity of the clusters of the special sector area 23 (or 22) of FIG. 1. Its damage often leads to the loss of all information on the HDD.

At the same time, damage to one sector in the data area 25 of FIG. 1 leads to destruction of only one information file.

If the clusters containing the file are not in a row, then the file data is fragmented. Heavily fragmented files significantly reduce the efficiency of the HDD-drive, since the read / write head, when searching for the next file cluster, must intermittently move between tracks and wait for the necessary sector to "approach".

Obviously, during the writing / reading of information, the magnetic heads are constantly shuttling between the special sector area, the transaction log area and the data sector area. For example, the magnetic head from the parking area 20 of FIG. 1 is shifted to the area of boot sectors 21, then to the area of special sectors 23 of FIG. 1, where NTFS files and tables are read, and in view of the bad cluster file data, free cylinder clusters that are registered in the bitmap file, the magnetic head is first moved to the transaction log area 25 of FIG. 1, where the planned changes are written to the log file, and then shifted to the data sector area 24 of FIG. 1, in the tracks of which the stored information is recorded. As a result, many microseconds are spent on writing / reading each sector of the HDD.

If a transaction has failed, for example, the power grid is disconnected, then using the records of the transaction log (log file), the HDD will automatically return to its original state.

Under certain circumstances, the boot sector area itself may become fragmented, which further reduces the performance of the HDD.

Shuttle mechanical movement of the magnetic heads of the HDD-drive between areas and the waiting for the necessary sector of each area to "fit" to the head, limit the overall speed of the computer as a single device.

At the same time, the area of special sectors, small in capacity, is the most frequently read and most frequently rewritten, which causes increased wear of the corresponding ferromagnetic tracks of the rotating physical disk. The destruction of the ferromagnetic surface in the area of special sectors leads to the irreversible loss of information of the entire disk.

For decades, magnetic recording methods have dominated the data storage segment. The HDD, invented in 1956, has become the bottleneck of the DCU due to its relatively slow speed.

--- "Solid State Drive" (hereinafter referred to as SSD-drive),

The problem in a computer or server with a magnetic HDD-drive is eliminated by replacing it with an SSD-drive (solid state drive), which provides much faster data processing speed.

The first samples had a capacity of 128 MB, but now they have become affordable models up to 480 GB.

SSD-drives are technologically complete read-only memory devices that store data in microcircuits. It differs from flash memory primarily in its large information capacity. The most common type of solid-state drives uses non-volatile memory such as NAND to store information, but there are options in which the drive is created on the basis of DRAM memory, equipped with an additional power source - a battery.

In order to optimize the use of SSD-drive in 2011, the logical interface NVMe (Non-Volatile Memory Express) was developed, support for which was added to Windows, starting with version 8.1.

Unlike an HDD, the price of an SSD is highly dependent on the available capacity, due to the limited density of memory cells and the limitation of the size of the crystal in the microcircuit.

Currently, SSDs are divided into four manufacturing technologies, which differ in storage density and resource:

• SLC (Single Level Cell) - one bit per cell

• MLC "Multi Level Cell" - two bits per cell

• TLC "Triple Level Cell" - three bits per cell

• QLC "Quad Level Cell" - 4 data bits per cell. The most budgetary drive with a cost of 1 TB is about $ 100-120.

Compared to HDDs, SSDs have the advantage of eliminating mechanical moving parts, hence:

• stability of the reading time of files, regardless of their location or fragmentation

• high read / write speed, up to several gigabytes per second for randomly located data, close to the throughput of interfaces (SAS / SATA II 300 MB / s, SAS / SATA III 600 MB / s)

• the number of random input-output operations per second (IOPS) of an SSD-drive is an order of magnitude higher than that of hard drives, due to the possibility of simultaneous launching of many operations and lower latency of each operation (there is no need to wait for a disk revolution before accessing, and also to wait pointing the disk head to the desired cylinder);

• much lower sensitivity to external electromagnetic fields;

• high mechanical resistance (withstand for a short time about 1500 g);

• small dimensions and weight. For solid-state drives, more compact standard sizes have been developed, for example, mSATA, NGFF (M.2).

• lack of noise;

• low power consumption;

But the main disadvantages of an SSD drive are considered several times higher cost per GB and a limited number of rewriting cycles, for example:

• Cheap drives (SB, SD, uSD type TLC) have a life on the order of 1000 cycles or less.

• Ordinary MLC, such as flash memory (flash memory), allows you to write data about 3000-10,000 times (guaranteed resource);

• More expensive types of SLC memory have a resource of the order of hundreds of thousands of rewriting cycles.

To combat uneven wear in high-performance (SATA and PCIe) SSDs, cell load balancing schemes are used: the main controller stores information about how many times which blocks have been overwritten, and, if necessary, writes to less worn blocks. When the real resource of memory banks is used up, the drive can switch to read-only mode, which allows you to save and copy data.

With the further rapid development of SSD technologies, the low resource in the number of rewrites will undoubtedly be eliminated, for example, by making an information carrier according to other physical principles, such as FeRam, ReRAM (resistive random-access memory), etc.

But now manufacturers of SSD-drives using 3D TLC technology give a guarantee of at least 3 years.

--- "Hybrid hard disk" (hereinafter referred to as SSHD-drive).

The world's first solid-state hybrid drive was developed by Seagate in 2007.

The SSHD is a combination of magnetic HDD and solid-state SSD parallel storage technologies. A separate solid state SSD with limited capacity is used as a large buffer (cache) of the information stored in the magnetic HDD. The device controller copies the most frequently used files from the data sector area of the HDD to the fast SSD.

At the heart of hybrid drive technology is deciding which data items are prioritized by an SSD and which are not. Therefore, SSHD drives can operate in two main modes:

• Automatic mode or self-optimized. In this mode, the hybrid hard drive independently makes all decisions related to the distribution of data and does not depend on the operating system

• Host-Optimized Mode or host-hinted. In this mode of operation, the SSHD hybrid drive includes an extended SATA command set (Hybrid Information)). Based on these commands, the operating system and device driver, taking into account the structure of the file system, decides which data items to place in the SSD.

Some specific features of an SSHD drive, such as host-hinted mode, require software support in the operating system. Support for host-hinted operation only appeared in Windows 8.1, while patches for the Linux kernel have been available since late 2014.

Due to the increased cache, it was possible to reduce the number of requests to the magnetic disk, which was reflected in a decrease in power consumption and heat dissipation, durability and a decrease in noise during operation. All this makes them an order of magnitude more productive and more practical than an HDD, but several times cheaper than an SSD drive.

As a result, the performance of the drive is increased. This combination allows you to take advantage (in terms of random IOPS) of an inexpensive SSD with a small capacity while maintaining the low cost of storing 1 GB of data on the HDD.

There are two main designs used to implement hybrid drives:

• with independent HDD and SSD enclosures, for example, Hyper Duo technology, Adaptec MaxIQ, each of which is connected to the computer via a separate interface cable

• HDD-drive, SSD-drive and controller are mounted in one case, for example, using Intel Smart Response technology. The case is connected to the computer with one interface cable. Structurally, the SSD is placed on the printed circuit board of the SSHD drive like a conventional microcircuit. We can say that this is one of the directions of the evolutionary development of HDD-drives.

The technical issue with SSHD hybrid drives is the tension between performance and cost. A noticeable performance gain occurs when copying all the most frequently used files from HDD to SSD. But for this it is necessary to increase the capacity of the SSD-drive, which increases the cost of the SSHD-drive and the cost of storing 1 GB of information. On the other hand, if there are no constantly used files, then intermediate copying of files from HDD to SSD will negatively affect the overall performance of the SSHD, including the number of input / output operations per second (IOPS). It does not make sense to install an SSD drive larger than 32 GB in a hybrid drive, since it will be cheaper to purchase a regular 64 GB SSD drive.

By the totality of the essential features of the invention, the closest analogue is the HDD-drive.

The main technical problem of hard disk drives is low performance. With a further increase in the capacity of the HDD-drive, the constant shuttle movement of the magnetic head between the tracks of the special sector area, the data sector area and the transaction log area significantly affects the number of random input-output operations per second (IOPS). The low read / write speed of the HDD has become a bottleneck in modern computers.

Disclosure of the essence of the invention

A magnetic storage device with divided areas (hereinafter NMRO) is a new type of CZU, in which the area of special sectors 8 Fig. 1 and the transaction log area 25 of FIG. 1 are removed from the ferromagnetic surface of the rotating physical disk 2 of FIG. 1 HDD to the corresponding SSD-mft 27 of FIG. 4 and SSD-log 29 of FIG. 4. The solid state drives 27 and 29 are located, for example, on the printed circuit board 10 of FIG. 4 NMRO controllers. Accordingly, the entire area of the ferromagnet of the rotating physical disk is occupied exclusively by the area of the data sectors 24 of FIG. 3. As a result:

• on the rotating physical disk 2 of FIG. 3 to the area of data sectors 24

fig. 3 the main amount of information is recorded, for example:

файлы операционной системы

operating system files

файлы драйверов, например, для работы с принтером или монитором о файл подкачки, который используется при недостаточности оперативной памяти

driver files, for example, for working with a printer or monitor o paging file, which is used when there is insufficient RAM

файлы прикладных программ

application files

другая информация для постоянного хранения

other information for permanent storage

• into a small SSD-mft 27 fig. 4, the area of special sectors of the NTFS system is written and stored, for example, the area of boot sectors, the MFT (Master File Table), bitmap file, bad cluster file, etc. file system NTFS

• into a small SSD-log 29 FIG. 4 writes and stores an NTFS transaction log area that contains a "log file" as well as copies of the boot sectors area, volume file, mirrored Master File Table, bad cluster file ... These files are simply saved, but are used when rolling back transactions or to recover the NMRO in the event of an SSD-mft crash.

Reading information from NMRO is as follows, for example:

• by the name of the file, the main controller 11 of FIG. 4 finds in the MFT table of the solid state memory SSD-mft 27 of FIG. 4 matching cluster numbers

• sectors of the clusters are sequentially read by the magnetic head 8 of FIG. 3 from the rotating physical disc 2 of FIG. 3

• information is sent directly to the central processor

• if necessary, portions of the read information are temporarily stored in the Cache memory (buffer memory) 12 of FIG. 4.

The recording of information in the NMRO is simplified as follows, for example:

• the file being written or part of it is stored in the memory cache 12 of FIG. 4

• according to the MFT table from the solid-state memory SSD-mft 27 fig. 4, the main controller 11 of FIG. 4 plans the cluster numbers on the rotating physical disc 2 of FIG. 3 to write a file to them

• the main controller 11 of FIG. 4 sequentially writes a file from the cache memory 12 of FIG. 4 to the scheduled clusters of the rotating physical disk 2 of FIG. 3

• in parallel, all planned changes on the rotating physical disk 2 of FIG. 3 are stored in the SSD-log 29 of FIG. 4

• if it is impossible to complete the transaction, the main controller 11 of FIG. 4, automatic restoration of the initial state of the clusters of the rotating physical disk 2 of FIG. 3 using "transaction log" entries in SSD-log 29 of FIG. 4

• After successful completion of the transaction, the modified MFT and bitmap file are copied from SSD-mft 27 of FIG. 4 in SSD-log 29 of FIG. 4

• in case of successful completion of the transaction and after a predetermined time has elapsed, the corresponding entry in the transaction log file log file of the SSD-log 29 of FIG. 4 is reset to zero.

NMRO retains a large information capacity and low cost of 1 GB HDD-drive, but compares favorably with it in higher reliability of information storage, less random access time, higher number of input-output operations per second (IOPS) and less mechanical wear of the ferromagnetic surfaces of the rotating physical disk 2 of FIG. 3, by eliminating the shuttle movement of the magnetic head 8 by the rocker arm 3 of FIG. 1 between the tracks of the NTFS special sector area 23 of FIG. 1, the tracks of the data sector area 24 of FIG. 1 and the tracks of the NTFS transaction log area 25 of FIG. 1 of the rotating physical disc 2 of FIG. 1, and also due to an order of magnitude higher read / write speed of the area of special sectors of the NTFS system in the SSD-mft 27 of FIG. 4 and the NTFS transaction log area in SSD-log 29 of FIG. 4 NMRO.

Brief Description of Drawings

FIG. 1 Hard disk drive (view without a cover from the side of a rotating physical disk), where 1 is a valve motor spindle, 2 is a rotating physical disk (non-removable hard disk covered with a ferromagnet), 3 is a rocker arm, 4 is an axis of a rocker arm, 5 is a voice coil rocker arms, 6 - permanent magnet, 7 - slider, 8 - magnetic head, 9 - preamplifier switch, 20 - magnetic head parking area, 21 - boot sector area, 22 - obsolete FAT32 special sector area, 23 - special sector area (modern NTFS), 24 - data sectors area, 25 - transaction log area.

FIG. 2 HDD controller, where 10 is the printed circuit board, 11 is the main controller, 12 is the cache memory (buffer memory), 14 is the valve motor control chip, 15-serial flash memory, 16 is the power connector, 17 is the connector for data cable connection.

Fig. 3 Magnetic drive with distributed areas (view without a cover from the side of a rotating physical disk), where 1 is a valve motor spindle, 2 is a rotating physical disk (non-removable hard, covered with a ferromagnet), 3 is a rocker arm, 4 is an axis of a rocker arm, 5 is rocker voice coil, 6 - permanent magnet, 7 - slider, 8 - magnetic head, 9 - switch-preamplifier, 20 - magnetic head parking area, 24 - data sectors area.

FIG. 4 NMRO controller, where 10 is a printed circuit board, 11 is the main controller, 12 is a cache memory (buffer memory), 14 is a valve motor control chip, 15 is a serial flash memory, 16 is a power connector, 17 is a connector for data cable connections, 27 - SSD-mft solid-state drive, 29 - SSD-log solid-state drive, 30 - connector (socket) for installing solid-state memory.

Implementation of the invention

In contrast to the HDD-drive on the printed circuit board of the NMRO controller, two SSD-drives of significantly small capacity and cost are additionally soldered.

After formatting the NMRO:

• over the entire area of the magnetic memory of the rotating physical disk 2 of FIG. 3, only the tracks of the data sector area 24 of FIG. 3

• in SSD-mft 27 fig. 4 the main table MFT and other files of the NTFS system are written

• in SSD-log 29 of FIG. 4 records the NTFS transaction log (log file), as well as copies of files from the SSD-mft, for example, the boot sector area, volume file, mirrored Master File Table, bad cluster file.

The low cost of SSD-mft and SSD-log is achieved due to their small capacity, for example 0.256 GB, as well as the use of the main controller 11, which is common with the rotating physical disk, in FIG. 4.

The high reliability of SSD-mft and SSD-log is achieved through the use of a cell load balancing scheme and the use of solid-state drives that allow a large number of cell rewrites, for example, using SLC technology.

The cell load balancing scheme sequentially blocks cells with a depleted resource of the number of rewrites. After almost complete use of free cells, NMRO is automatically transferred to the "read-only" mode, which ensures the safety and availability of information on the rotating physical disk 2 FIG. 3.

SSD-mft and SSD-log are housed in removable enclosures such as micro-SD (formerly known as TransFlash, T-Flash or TF), which are inserted into the respective Socket 30 of FIG. 4 printed circuit boards 10, which allows you to transfer solid-state drives 27 and / or 29, which have exhausted their resource in the number of rewrites, to a separate stand and prophylactically copy information from SSD-mft and / or SSD-log to new SSD-drives. New SSD-mft and / or SSD-log with the transferred data are inserted into the appropriate NMRO sockets and its work continues without loss of information.

Moreover, the location of the connector (socket) 30 of FIG. 4 on a printed circuit board allows the replacement of SSD-mft 27 and / or SSD-log 29 without disassembling the NMRO case and the computer case, for example, as in a cell phone.

The easily removable SSD-mft 27 and SSD-log 29 have an electronic security key function. After removing them from the socket of the NMRO controller, the data sector area 24 of FIG. 3 on the rotating physical disc 2 of FIG. 3 is a collection of sectors of actual and "remote" information with an admixture of operating system file sectors, which are comparable to tiny pieces of documents shredded in an office shredder. Moreover, unlike the password, the removal of the SSD-mft and SSD-log of the active disk from the NMRO makes the computer's operating system absolutely inaccessible, for example, to hackers. During uncontrolled transportation of a computer, for example, in the luggage of an aircraft, seized SSD-mft and SSD-log are moved through an independent channel, for example, by a computer user or courier.

Claims (22)

1. A magnetic drive with divided areas (hereinafter referred to as the DRIVE) (or briefly: HMPO, MDDA ("Magnetic Device with Divided Areas")), which includes an HDD (Hard (magnetic) Disk Drive), in which a rotating ferromagnetic layer of the physical disk by the magnetic head is sequentially recorded with a breakdown by cylinders and sectors, in particular, the area of data sectors with files of the operating system, drivers and application programs, and the main controller is located on the PCB of the HDD controller, which controls the process of writing / reading / formatting on a rotating physical disk, as well as microcircuits of cache memory (buffer memory), control of a valve motor, serial flash memory with the program of the main controller, a connector for connecting a power supply and a connector for connecting an information cable, characterized in that two non-volatile solid state drives (solid state drives) with a load balancing scheme are additionally soldered to the PCB of the HDD controller, hereinafter referred to as SSD-mft and SSD-log drives, in combination with which the printed circuit board is referred to hereinafter: "controller NMRO ", and the main controller on the printed circuit board of the NMRO controller controls a rotating physical disk, an SSD-mft drive and an SSD-log drive, where the NMRO's SSD-mft drive is intended for recording and storing NTFS files, including: - boot sector areas; - an area of special sectors, which contains, for example: - volume file; - root directory; - file with MFT-table (Master File Table) of the NTFS operating system, where the name of the files and their distribution by cylinder numbers, magnetic heads and sectors on the tracks are recorded; - a mirror copy of the file with the MFT table; - bitmap file; - bad cluster file; - file of attributes (attribute definition table), and the SSD-log drive of the NMRO controller is intended for recording and storing: - the area of the transaction log of the NTFS system, which contains the transaction log (log file); - as well as copies of SSD-mft files of the NMRO controller, including: - boot sector areas; - volume file; - mirrored Master File Table; - bad cluster file, the area of the data sectors occupies all the surfaces of the ferromagnetic layers of the rotating physical disk. 2. STORAGE DRIVE according to claim 1, characterized in that for the purpose of, for example, preventive maintenance, which provides for periodic transfer of files and tables (without loss of information by the STORAGE DRIVE) from the exhausted SSD-mft and / or SSD-log to new SSD- drives, SSD-mft and / or SSD-log are made in a removable small-sized case, for example micro-CD, each of which is inserted into the corresponding connector of the NMRO controller PCB, where the DRIVE design allows replacing the SSD-mft and / or SSD-log without it disassembly. 3. The STORAGE DRIVE according to claim 1, characterized in that for the purpose, for example, of protecting the DRIVE's information during assembly, transportation and / or storage of the computer, removable SSD-mft and / or SSD-log can function as electronic keys that can be stored, for example , in a safe or transported separately from the computer, for which the design of the DRIVE and the computer provide removal / installation of SSD-mft and / or SSD-log without disassembling the computer, such as replacing the micro-CD in a smartphone.

Images