Storage system, host system, and storage system operation method
By dividing high-reliability blocks and spare blocks in the storage system, the problem of improper metadata management in open channel SSDs is solved, data reliability and performance are improved, and the rapid error repair of metadata is achieved.
Patent Information
- Application Number
- CN202010745090.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-07-29
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2040-07-29
AI Technical Summary
In an open channel SSD environment, it is difficult for the host system to accurately and timely determine the block characteristics in the storage system, resulting in improper metadata management and affecting data reliability.
By dividing high-reliability blocks and spare blocks in the storage system, the controller provides block information to the host system, which assigns high-reliability blocks to store metadata and performs error repair operations in the high-reliability block. , ensure the reliability of metadata.
Improves the reliability and overall performance of metadata, reduces system management overhead, and quickly fixes errors when block replacement, ensuring the reliability of data processing.
Smart Images

Figure CN112597069B_ABST
Abstract
Description
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS
[0002] This application claims the benefit of Korean Patent Application No. 10-2019-0122654 filed on October 2, 2019, in the Korean Intellectual Property Office, the subject matter of which is incorporated herein by reference. Technical Field
[0003] The inventive concept relates to a storage system, and more particularly, to a storage system capable of effectively managing metadata. The inventive concept also relates to a host system for controlling such a storage system and a storage system operating method. Background Art
[0004] A solid state drive (SSD) is an example of a storage system based on a semiconductor storage device such as flash memory. Various interfaces have been used to access storage systems such as SSDs (e.g., Serial Advanced Technology Attachment (SATA), Peripheral Component Interconnect (PCI) Express (PCIe), and Serial Attached Small Computer System Interface (SCSI) (SAS), Non-Volatile Memory Express (NVMe), etc.).
[0005] Open channel SSD (OC-SSD) has recently been proposed. In an open channel SSD, various functions such as address translation operations that were once performed in an SSD are performed in a host system. As a result, the host system includes a flash translation layer (FTL) that was once managed by the SSD. Therefore, the host system can directly perform various management operations based on the FTL. However, before receiving information about bad blocks, it may be difficult for the host system to determine one or more block characteristics of the storage system. Such bad block information may be transmitted via a separate command, and because of this, it may be difficult to maintain data reliability because the host system has to manage important information such as metadata without accurately knowing the block characteristics. Summary of the invention
[0006] The present inventive concept provides a storage system, a host system, and an operating method for the storage system, which effectively manage metadata to enhance the reliability of the metadata.
[0007] According to one aspect of the inventive concept, there is provided a storage system that communicates with a host system, the storage system comprising: a storage device, the storage device comprising a storage medium, the storage medium being divided into a plurality of blocks including a high reliability block and a spare block; and a controller. The controller: provides block information identifying the high reliability block among the plurality of blocks to the host system; receives a block allocation request from the host system, wherein the block allocation request is defined with reference to the block information and identifies at least one high reliability block to be used for storing metadata; and allocates at least one high reliability block to a meta-area in response to the block allocation request. The controller comprises: a bad block manager that manages a block allocation operation performed in response to the block allocation request, and repairs an error in metadata stored in one of the high reliability blocks.
[0008] According to another aspect of the inventive concept, there is provided an operation method of a storage system including a plurality of blocks. The operation method comprises: receiving a block allocation request from a host system, the block allocation request identifying a high reliability block among the plurality of blocks to be used for storing metadata; in response to the block allocation request, allocating the high reliability block and at least one corresponding spare block to a meta area; in response to a metadata write request from the host system, writing the metadata into the high reliability block, writing the metadata into the at least one corresponding spare block, and writing repair information for repairing errors in the metadata into the high reliability block; determining whether one of the high reliability blocks is a bad block; and transmitting to the host system one of the metadata read from the high reliability block and the error-repaired metadata generated using the repair parity information.
[0009] According to another aspect of the inventive concept, there is provided a host system communicating with a storage system including a plurality of blocks. The host system includes: a host flash translation layer (FTL) used in implementing management operations on the plurality of blocks, the host FTL including a request generation module, the request generation module generating a block allocation request, the block allocation request being used to allocate a high reliability block to be used to store metadata among the plurality of blocks; and an interface circuit connected to the storage system through an interface to receive block information identifying the high reliability block, and transmitting the block allocation request from the host system to the storage system, wherein the block allocation request includes information identifying the number and corresponding positions of the high reliability blocks. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Embodiments of the present invention will be more clearly understood through the following detailed description in conjunction with the accompanying drawings, in which:
[0011] Figure 1is a block diagram showing a data processing system according to an embodiment;
[0012] Figure 2 is a block diagram showing an example of blocks allocated to a meta area and a user area;
[0013] Figure 3 is a conceptual diagram illustrating an operation method of a data processing system according to an embodiment;
[0014] Figure 4 is a block diagram showing an implementation example of a data processing system according to an embodiment;
[0015] Figure 5 is a block diagram showing an implementation example of a controller included in a storage system according to an embodiment;
[0016] Fig. 6A and Figure 6B is a block diagram illustrating an example of repairing an error occurring in metadata according to an embodiment;
[0017] Figure 7 is a block diagram illustrating an implementation example of a host flash translation layer (FTL) included in a host system;
[0018] Figure 8 is a flowchart illustrating an operating method of a data processing system according to an embodiment;
[0019] Fig. 9A and Fig. 9B is a conceptual diagram showing an operation example of the data processing system according to the embodiment;
[0020] Fig.10 is a block diagram showing a data processing system according to a modifiable embodiment;
[0021] Fig.11A and Fig. 11B is a block diagram illustrating various examples of block allocation operations performed in a storage system;
[0022] Fig. 12A and Fig. 12B is a block diagram illustrating an example of allocating blocks to a meta region in a solid state drive (SSD) including a plurality of storage media; and
[0023] Fig.13 is a block diagram illustrating a network system including a server system according to an embodiment. DETAILED DESCRIPTION
[0024] Hereinafter, embodiments of the inventive concept will be described in more detail with reference to the accompanying drawings.
[0025] Figure 1 is a block diagram illustrating a data processing system 10 according to an embodiment.
[0026] Reference Figure 1 , the data processing system 10 may include a host system 100 and a memory system 200, wherein the memory system 200 may include a controller 210 and a memory device 220. The host system 100 may provide a request for a read, write, and / or erase operation to the memory system 200. And, in response to such a request received from the host system 100, the memory system 200 may perform a read operation, a write operation, and / or an erase operation on data specified by one or more addresses associated with the request.
[0027] The storage system 200 may include a plurality of storage media configured to store data in response to a request received from the host system 100. For example, the storage system 200 may include one or more solid-state drives (SSDs). When the storage system 200 includes an SSD, the storage device 220 may include a plurality of flash memory chips (e.g., NAND flash memory chips) each configured to store data non-volatilely. Alternatively, the storage device 220 may correspond to one flash memory device, or the storage device 200 may include a memory card having one or more flash memory chips.
[0028] When the storage system 200 includes flash memory, the flash memory may include a two-dimensional (2D) NAND memory array or a three-dimensional (3D) (or vertical) NAND (VNAND) memory array. A 3D memory array may be a circuit including an active area disposed on a silicon substrate and associated with the operation of each memory cell, and may be configured in a monolithic type at the physical level of at least one memory cell array, each memory cell array including circuits disposed on or in a substrate. The monolithic type may mean that the layers of each level constituting the array are stacked directly on the layers of the lower levels of the array.
[0029] In an embodiment, a 3D memory array may include a plurality of vertical NAND strings arranged in a vertical direction so as to place at least one memory cell on another memory cell. At least one memory cell may include a charge trapping layer.
[0030] In this regard, U.S. Patent Publication Nos. 7,679,133, 8,553,466, 8,654,587, and 8,559,235 and U.S. Patent Application No. 2011 / 0233648 disclose suitable elements related to certain 3D memory cell arrays that include multiple levels and in which word lines and / or bit lines are shared between the multiple levels. The common subject matter of these documents is incorporated herein by reference.
[0031] As another example, the memory system 200 may include various memories. For example, the memory system 200 may include a non-volatile memory, where examples of the non-volatile memory may include, for example, a magnetic random access memory (RAM) (MRAM), a spin-transfer torque MRAM, a conductive bridge RAM (CBRAM), a ferroelectric RAM (FeRAM), a phase change RAM (PRAM), a resistive RAM (ReRAM), a nanotube RAM, a polymer RAM (PoRAM), a nano floating gate memory (NFGM), a holographic memory, a molecular electronic memory, and an insulator resistive memory.
[0032] The host system 100 may communicate with the storage system 200 using one or more interfaces. According to an embodiment, the host system 100 may communicate with the storage system 200 via one or more interfaces (e.g., Universal Serial Bus (USB), Multimedia Card (MMC), Peripheral Component Interconnect (PCI) Express (PIC-E), Advanced Technology Attachment (ATA), Serial ATA (SATA), Parallel ATA (PATA), Small Computer System Interface (SCSI), Serial Attached SCSI (SAS), Enhanced Small Disk Interface (ESDI), and Integrated Drive Electronics (IDE)). In this regard, a non-volatile memory express (NVMe) interface has been proposed as an interface optimized for the storage system 200 such as an SSD, and according to an embodiment, NVMe may be applied to the data processing system 10.
[0033] As an implementation example of the storage system 200, the storage system 200 may include an open channel SSD (hereinafter referred to as "OC-SSD"). In the OC-SSD environment, the host flash translation layer (FTL) 110 may be included in the host system 100, and thus, various management operations corresponding to the storage system 200 may be performed under the control of the host system 100. For example, the host system 100 may perform an address conversion operation between a logical address (LA) recognized by the host system and a corresponding physical address (PA) recognized by the storage system 200. Therefore, the host system 100 may provide a physical address to the storage system 200, wherein such an address may be used to directly index or identify data associated with a host system request.
[0034] according to Figure 1 In the illustrated embodiment, in addition to the host FTL 110, the host system 100 may further include a first interface circuit 120. The controller 210 of the storage system 200 may include a bad block manager 211 and a second interface circuit 212. The storage device 220 may include a plurality of storage media, wherein each storage medium may include one or more storage devices (e.g., flash memory chips). In addition, each storage medium may include a storage cell array 221 having a plurality of blocks. Figure 1 , it can be assumed that one or more memory cell arrays 221 provided by the memory device 220 are included in one storage medium.
[0035] In order to reliably perform data processing over an extended period of time, the storage system 200 must perform various management operations. Examples of commonly performed management operations include: LA to PA and / or PA to LA address translation, block mapping including bad block mapping, and block management, etc. Many management operations may require access to (or be based on) certain metadata stored by the storage system 200. In this case, the reliability and timely access to the stored metadata are important factors in the successful operation of the storage system 200.
[0036] However, in some conventional OC-SSD environments, the host system 100 may be unable to accurately and timely determine one or more characteristics of the data stored in the storage device 200 in some cases. For example, the host system may not know the physical characteristics associated with bad blocks in the storage device 220. Therefore, the host system 100 may be substantially limited in its management of key information (e.g., metadata).
[0037] For example, when it is determined that a particular block (e.g., a first block) of the storage device 220 is a bad block, the host system 100 may replace (or map) the first block to another block (e.g., a second block) known to be a normal block. Here, the terms "bad" and "normal" are relative in nature and may be defined differently. However, in the description of certain embodiments, a bad block is a block that is considered to be unsuitable for storage of metadata in its characteristics. Once the first block address has been mapped to the second block, the resulting (or updated) block mapping information may be used to prevent critical data (e.g., metadata) from being physically stored in the first block.
[0038] Unfortunately, when a first block is used to store metadata and then is determined to be a bad block, block mapping information associated with the first block is mapped to a second block, and metadata is stored in the second block, the host system 100 may have problems when it cannot check the nature and / or location of the second block now identified by the block mapping information. That is, when the data processing system 10 is powered on again after being powered off, the host system 100 must read the metadata from the storage device 220 and perform certain management operations associated with the stored metadata. However, if the host system cannot accurately or timely know that the specific block storing the metadata has been physically changed from the first block to the second block, this process may not proceed as expected.
[0039] According to some embodiments of the inventive concept, in order to ensure the reliability of key information such as metadata and perform effective management, the storage system 200 may perform some management operations that can be applied to the data processing system 10. According to an embodiment, the bad block manager 211 may perform management operations associated with bad blocks, for example, the bad block manager 211 may perform bad block management operations on blocks storing key information such as metadata. In addition, in order to ensure the reliability or integrity of metadata, the bad block manager 211 may perform repair operations on metadata stored in high reliability blocks.
[0040] The host system 100 and the storage system 200 can use the first interface circuit 120 and the second interface circuit 212 to transmit and / or receive (hereinafter referred to as "transmit") requests, data and / or information in various manners while operating in compliance with at least one protocol and data format (e.g., the protocol and data format defined for the NVMe interface).
[0041] The data storage space (or capacity) provided by the storage device 220 of the storage system 200 may be physically and / or conceptually divided into a plurality of blocks. The blocks may be identified by specific designation and / or location within the storage device 220. Among the plurality of blocks, some blocks may be designated as providing high data storage reliability (hereinafter referred to as "high reliability blocks"). The data reliability of a particular block may be determined in various ways, and the designation of a high reliability block is relative. For example, a high reliability block may be a so-called error-free (EF) block that exhibits relatively few errors (or no errors) during testing, or is calculated to provide a very low relative error probability.
[0042] However, when the “block information” Info_BLK is transmitted from the storage system 200 to the host system 100, the storage system 200 may identify the determined or designated plurality of high reliability blocks to the host system. Referring to the block information Info_BLK, the host system 100 may transmit a “block allocation request” Req_BLK to the storage system 200. Here, the block allocation request Req_BLK is used to allocate a high reliability block for storing critical data (e.g., metadata) with reference to the block information Info_BLK.
[0043] The block information Info_BLK may include various information. For example, the block information Info_BLK may include information identifying (by location and / or designation, such as a block address) one or more high reliability blocks within the storage system 200. The block information Info_BLK may also include information associated with repair techniques (or repair operations) that may be used to support or improve data integrity of certain blocks.
[0044] As described above, high reliability blocks in the storage system 200 may be determined in various ways. For example, some of the multiple blocks provided by the storage device 220 may include a multi-level memory cell (MLC) (i.e., a memory cell that is generally used to store two or more data bits per memory cell), wherein the MLC is instead used to store only a single bit of data. In other words, some of the high reliability blocks will be blocks that include an MLC used as a single level memory cell (SLC) to store a single bit of data. Alternatively, the storage system 200 may perform a self-test operation to determine a block that provides relatively high reliability, or a process for selectively enhancing blocks at specific locations may be applied during a manufacturing process, thereby providing high reliability blocks in the storage system 200.
[0045] Additionally or alternatively, the storage system 200 may perform repair operations to improve data reliability during writing metadata to a block and / or reading metadata from a block. For example, various data recovery methods (such as redundant array of inexpensive disks (RAID) or data mirroring) may be applied to the storage system 200.
[0046] The block allocation request Req_BLK transmitted from the host system 100 to the storage system 200 may be defined differently to include various information. For example, the host system 100 may use the block allocation request Req_BLK to indicate the number and corresponding positions of high reliability blocks to be used as blocks for storing critical data such as metadata. Therefore, the block allocation request Req_BLK may include information identifying the number and / or position of high reliability blocks. The block allocation request Req_BLK may also be used to identify a repair operation to be applied to the data processing system 10 from a number of available repair operations (or techniques).
[0047] The data storage space (or capacity) provided by the storage device 220 of the storage system 200 may be further divided (physically and / or conceptually) into a plurality of regions, wherein each region includes one or more blocks. A region including one or more high-reliability blocks may be designated as a "meta region", while a region including only blocks storing non-metadata (e.g., user data) may be designated as a "normal region". In this regard, certain blocks provided by the storage device 220 may be designated as "spare blocks". The spare blocks may be unallocated (or not currently used to store data), and may be normal blocks in a normal region and / or high-reliability blocks in a meta region.
[0048] exist Figure 1 In the illustrated embodiment, the storage device 220 provides a meta region Reg_Meta including “M” blocks BLM1_1 to BLK1_M and a normal region Reg_Normal including “N” blocks BLM2_1 to BLK2_N, where “M” and “N” are integers greater than 1.
[0049] With this exemplary configuration, the host system 100 can identify (or select) a repair operation and one or more high reliability spare blocks included in the meta area Reg_Meta used when performing the repair operation. The allocation of high reliability spare blocks can be performed in various ways. For example, the host system 100 can select one or more spare blocks, and information identifying the number and / or position of the spare blocks can be included in the block allocation request Req_BLK. Or, information associated with the selected repair operation can be included in the block allocation request Req_BLK. Therefore, the data processing system 10 can be implemented so that the storage system 200 can specify one or more high reliability spare blocks in the meta area Reg_Meta in response to the block allocation request Req_BLK. In this regard, it should be noted that, depending on the nature of the selected repair operation, different numbers of spare blocks may be required. Therefore, when the host system 100 and / or the storage system 200 allocate one or more blocks as spare blocks, the selected repair operation must be considered.
[0050] Therefore, the host system 100 may provide the identification (e.g., address) of one or more high reliability blocks in the meta area Reg_Meta for storing critical data such as metadata as part of the block allocation request Req_BLK to the storage system 200. In contrast, when the host system 10 issues a block allocation request Req_BLK for one or more normal blocks in the normal area Reg_Normal storing user data, the host system 100 may provide the storage system 200 with addresses for identifying the normal blocks.
[0051] According to an embodiment, in an OC-SSD environment, the storage system 200 may provide information about high reliability blocks to the host system 100, and the host system 100 may select one or more high reliability blocks to store critical data such as metadata, thereby enhancing or ensuring the integrity of the metadata. The system management overhead performed by the host system 100 may be reduced, thereby improving the overall performance of the data processing system 10. Moreover, even when it is determined that the block storing metadata is a bad block, one or more errors in the metadata may be easily repaired by using the above-mentioned method of performing the selected repair operation. In addition, even when the block storing metadata is replaced by another block, the host system 100 may access the information stored in the high reliability block included in the meta area Reg_Meta, or may access only the information stored in the high reliability spare block corresponding to the bad block, thereby checking the block mapping information. Therefore, the time required to check the information associated with the metadata may be reduced, and the overall efficiency of the management of critical data such as metadata may be enhanced.
[0052] Figure 2is a block diagram further illustrating a method for allocating blocks in a meta region Reg_Meta and a normal region Reg_Normal in one example. Here, Figure 2 The illustrated example assumes the use of NAND flash memory as the storage medium.
[0053] Reference Figure 1 and Figure 2 , the storage cell array 221 may include a plurality of blocks including one or more high reliability blocks. The storage system 200 may transmit the block information Info_BLK to the host system 100, and the host system 100 may transmit a block allocation request Req_BLK that may be used to select at least one high reliability block for storing metadata. In addition, the host system 100 may select a repair operation that may be used to repair erroneous metadata, and select at least one spare block Rsvd that may be used to store information during the repair operation.
[0054] exist Figure 2 In the example shown in FIG. 4 , four (4) high reliability blocks (EF) and four (4) spare blocks are allocated to the meta region Reg_Meta. In other embodiments, a different number of high reliability blocks and / or spare blocks may be allocated to the meta region Reg_Meta.
[0055] Various operations for repairing errors in metadata may be applied, and according to various embodiments, operations using data mirroring or repairing parity may be applied. For example, in the case of applying a data mirroring operation, the host system 100 may issue a request to write metadata to a high reliability block, and the storage system 200 may write the metadata to the high reliability block and a spare block corresponding thereto. Subsequently, when a high reliability block storing metadata is determined to be a bad block, the storage system 200 may map the address of the high reliability block to a corresponding spare block, and may store block mapping information therein. In addition, when the host system 100 issues a request to access metadata stored in a high reliability block, the storage system 200 may access the corresponding spare block with reference to the block mapping information.
[0056] According to an embodiment, in the case where the repair operation includes using repair parity information, the host system 100 may issue a request to write metadata to a high reliability block, and the storage system 200 may generate repair parity information from the metadata of the write request and / or other metadata constituting one repair unit, and may store the generated repair parity information to a spare block. For example, the other metadata constituting the repair unit may be stored in a block different from the high reliability block of the write request (e.g., at least another high reliability block). For example, when a high reliability block storing metadata requested to be read by the host system 100 is determined to be a bad block, the storage system 200 may repair errors in the metadata based on an arithmetic operation using the repair parity information and other metadata of the repair unit, and may provide the metadata with the repaired errors to the host system 100.
[0057] In addition, the correctable error rate can be changed by adjusting the number of spare blocks included in the meta region Reg_Meta or by adjusting the ratio of the number of spare blocks to the number of high reliability blocks. For example, when there are many spare blocks, relatively more parity bits can be stored based on the same bit of data, and therefore, the error correction capability can be enhanced. For example, when there are many spare blocks, an error correction circuit with high error correction capability can be applied to the storage system 200, in which case the ability to correct errors in metadata can be further enhanced.
[0058] exist Figure 2 In the embodiment of the invention, a data mirroring method in which metadata is copied once has been described, but the embodiment is not limited thereto. For example, metadata may be copied twice (or more times) and may be stored in a spare block, and when many spare blocks are allocated, a repair operation based on multiple copy operations may be applied.
[0059] Figure 3 is shown in an example Figure 1 A conceptual diagram of a method of operating the data processing system 10. Figure 3 In the present invention, a flash memory-based OC-SSD is assumed as an example of a storage system, and therefore, a host system HOST is assumed to include a flash translation layer (FTL).
[0060] Once the OC-SSD is connected to the host system HOST, a device initialization process may be performed, wherein metadata required for managing the OC-SSD must be stored during the device initialization process. To this end, the OS-SSD may transmit (e.g., send) block information Info_BLK that describes a meta area including one or more blocks that may have been set during the device formatting process. And, as described above, the block information Info_BLK provided by the OC-SSD to the host system HOST may include information describing one or more high reliability blocks in the meta area Reg_Meta, one or more spare blocks, and one or more recovery operations supported by the OC-SSD.
[0061] Upon receiving the block information Info_BLK, the host system HOST may issue a block allocation request Req_BLK to the meta area Reg_Meta requesting allocation of one or more high reliability blocks. For example, the number and corresponding positions of the high reliability blocks to be allocated to the meta area may be selected by the host system HOST, and information identifying at least one recovery operation to be applied to the meta area may be included in the block allocation request Req_BLK. In response to the block allocation request Req_BLK from the host system HOST, the OC-SSD may allocate one or more high reliability blocks to the meta area. The OC-SSD may also allocate one or more spare blocks to be used during a repair operation to the meta area. The OC-SSD may then transmit information indicating that the requested allocation has been completed to the host system HOST.
[0062] Subsequently, when a data access operation is performed by the OC-SSD, metadata related to the data access operation may be stored or updated in the meta area, and various repair operations (e.g., data mirroring and recovery algorithms) may be used to correct or repair errors in the metadata. In addition, the OC-SSD may perform a bad block management operation on a block assigned to the meta area, and may store block mapping information generated during the management of the bad blocks. According to an embodiment, the OC-SSD may transmit "bad block information" associated with the management of the bad blocks to the host system HOST, and the host system HOST may then accurately determine the current block state of the meta area for storing metadata in response to the bad block information.
[0063] Figure 4 is a block diagram illustrating a data processing system 300 in one example in which a meta region associated with each of a plurality of NAND flash memory devices is used to store critical data such as metadata.
[0064] Reference Figure 4 , the data processing system 300 may include a host system 310 and a storage system 320. Figure 4In the embodiment, for ease of description, only the first flash memory array NAND1 321 and the second NAND flash memory array NAND2 322 (i.e., the first storage medium and the second storage medium) are shown in the storage system 320. However, those skilled in the art will recognize that the storage system 320 may also include additional storage media and control and logic components for performing at least management operations associated with blocks for storing critical data such as metadata.
[0065] The host system 310 may include a management module (e.g., a NAND management module, not shown) capable of performing certain management operations associated with the storage system 320. Thus, the management module may include an FTL (or host FTL). Figure 4 In the embodiment, the first host FTL 311 can be used to manage the first storage medium 321, and the second host FTL 312 can be used to separately manage the second storage medium 322. However, a single FTL can be used to manage multiple storage media. In addition, according to the above embodiment, the storage system 320 can transmit the block information Info_BLK to the host system 310, and the host system 310 can transmit the block allocation request Req_BLK to the storage system 320. For example, the corresponding block information Info_BLK and the block allocation request Req_BLK can be transmitted for each NAND flash memory device, wherein the first host FTL 311 transmits the first block allocation request Req_BLK to the first NAND flash memory array 321 / the first NAND flash memory array 321 transmits the first block information Info_BLK to the first host FTL 311, and the second host FTL 312 transmits the second block allocation request Req_BLK to the second NAND flash memory array 322 / the second NAND flash memory array 322 transmits the second block information Info_BLK to the second host FTL 312. Alternatively, a single block information Info_BLK may be used to describe the memory system 320 to the host system 310 , and a single block allocation request Req_BLK may be used to allocate blocks in the memory system 320 in various ways.
[0066] It is assumed that each of the first flash memory array 321 and the second flash memory array 322 includes a plurality of blocks, wherein each of the plurality of blocks includes at least one high reliability block (EF) and at least one spare block (Rsvd). Under this assumption, based on the first block allocation request Req_BLK from the host system 310, meta areas 321_1 and 322_1 may be respectively set in the first storage medium 321 and the second storage medium 322, and one or more high reliability blocks and spare blocks Rsvd may be allocated to each of the meta areas 321_1 and 322_1. Figure 4, an example in which various blocks are allocated differently on the meta area 321_1 of the first storage medium 321 and the meta area 322_1 of the second storage medium 322 is shown.
[0067] According to an embodiment, first metadata associated with the first storage medium 321 may be stored in a meta area 321_1 of the first storage medium 321, and second metadata associated with the second storage medium 322 may be stored in a meta area 322_1 of the second storage medium 322. It is further assumed that a repair operation using a data mirroring technique may be applied to the meta area 321_1 of the first storage medium 321. Therefore, the number of spare blocks is equal to the number of high reliability blocks allocated in the meta area 321_1. On the other hand, it is assumed that a repair operation using a repair parity technique may be applied to the meta area 322_1 of the second storage medium 322. Therefore, a smaller number of spare blocks than high reliability blocks may be allocated in the meta area 322_1.
[0068] Figure 5 is a block diagram illustrating, in one example, a controller 400 that may be incorporated in a storage system according to an embodiment.
[0069] Reference Figure 5 , the controller 400 may include a central processing unit (CPU) 410 or processor, a host interface circuit 420, a memory interface circuit 430, and a working memory 440. Here, the controller 400 may further include an error correction code (ECC) engine 450.
[0070] The CPU 410 may execute various applications (e.g., programs) stored in the working memory 440 to control the overall operation of the storage system. The host interface circuit 420 may control communication with the host system according to an interface using at least one established data format and at least one communication protocol (e.g., a protocol consistent with NVMe). In addition, the memory interface circuit 430 may provide an interface with a plurality of storage media included in the storage device, for example, the memory interface circuit 430 may perform independent communication with the storage medium through a plurality of channels. In addition, the ECC engine 450 may perform operations for detecting and correcting errors in data stored in the storage device, for example, an ECC parity may be generated from written data, and an error detection operation and an error correction operation may be performed by using the read data and the ECC parity corresponding thereto.
[0071] In addition, programs for performing various functions according to embodiments and functions of the storage system may be loaded into the working memory 440. The working memory 440 may be implemented as various types such as RAM, read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other memories.
[0072] According to an embodiment, the meta-area management module 441, the bad block management module 442, and the repair module 443 may be loaded into the working memory 440. For example, according to the above-mentioned embodiment, the host interface circuit 420 may communicate with the host system to receive various requests from the host system, transmit various information, and receive a block allocation request requesting allocation of one or more high reliability blocks to store metadata. The CPU 410 may execute the meta-area management module 441, and based on the block allocation request from the host system, may perform an operation of allocating one or more high reliability blocks and / or spare blocks to the meta-area.
[0073] Metadata (as an example of key data) may be stored in a high reliability block and / or a spare block included in the meta area, and based on the execution of the bad block management module 442, a management operation may be performed on a bad block identified in the meta area. For example, when a high reliability block (e.g., a first high reliability block) is determined to be a bad block, the bad block management module 442 may manage a block mapping to replace the first high reliability block with another high reliability block. In addition, the bad block management module 442 may perform a block management operation indicating a spare block storing repair parity information or a high reliability block storing other metadata constituting the same repair unit as the metadata of the first high reliability block. In addition, by executing the repair module 443, an error occurring in the first high reliability block may be repaired, for example, an access operation may be performed on a spare block corresponding to the first high reliability block, or the error may be repaired by an arithmetic operation using the repair parity and other metadata constituting the repair unit.
[0074] exist Figure 5 In FIG. 1 , an example of performing the metadata management operation and the bad block management operation according to the embodiment by software is shown, but the embodiment is not limited thereto. Figure 5 In the embodiment of the present invention, at least some of the operations performed by executing software loaded into the working memory 440 may be performed by using a circuit implemented in hardware.
[0075] Fig. 6A and Figure 6B 1 and 2 are conceptual block diagrams respectively illustrating, in one example, repairing an error occurring in metadata according to an embodiment.
[0076] Reference Fig. 6A, a first example is shown, in which the first high reliability block EF 1 to the fourth high reliability block EF 4 storing metadata and the corresponding first spare blocks Rsvd 1 to the fourth spare blocks Rsvd 4 are allocated to the meta area. The host system can issue a request to write metadata to the first high reliability block EF 1 to the fourth high reliability block EF 4, and as part of the data mirroring technology, the storage system can write the metadata written to the first high reliability block EF 1 to the fourth high reliability block EF 4 to the first spare blocks Rsvd 1 to the fourth spare blocks Rsvd 4.
[0077] For example, the host system may provide the storage system with the first address ADD11 indicating the first high reliability block EF 1 and metadata, and the storage system may write the metadata into the first high reliability block EF 1, and may also write the metadata into the first spare block Rsvd 1 corresponding to the first high reliability block EF 1. The storage system may know the second address ADD21 indicating the first spare block Rsvd 1, and may internally manage block mapping information including mapping information about the first address ADD11 and the second address ADD21. Subsequently, in the case where the first high reliability block EF 1 is determined to be a bad block, when a request to access the metadata of the first high reliability block EF 1 is received from the host system, the storage system may write the metadata into the first spare block Rsvd 1, or may read the metadata from the first spare block Rsvd 1.
[0078] In addition, the metadata part may include repair information (or repair units), and repair parity information corresponding to the repair units may be generated and stored in the meta area. The storage system may perform a RAID-based data recovery algorithm by using the metadata part, and may store the resulting repair parity information generated based on the result of performing the recovery algorithm in the spare block. Figure 6B , another example is shown, in which the first metadata Meta DATA 1 to the fourth metadata Meta DATA 4 are stored in the first high reliability block EF 1 to the fourth high reliability block EF 4, a repair unit is configured and the corresponding repair parity information is stored in the first spare block Rsvd 1.
[0079] The repair parity information must be updated when data of at least one of the first metadata MetaDATA 1 to the fourth metadata Meta DATA 4 respectively stored in the first high reliability block EF 1 to the fourth high reliability block EF 4 is changed. Therefore, when the first high reliability block EF 1 is determined to be a bad block, the error in the first metadata Meta DATA 1 can be repaired using the second metadata Meta DATA 2 to the fourth metadata Meta DATA 4 respectively stored in the second high reliability block EF 2 to the fourth high reliability block EF 4 as the repair parity information.
[0080] Figure 7 is a block diagram showing a host FTL 500 that can be used as a host system in one example. Figure 7, the host FTL 500 may be loaded into a working memory (not shown) of the host system, and based on the functions implemented by the software, the host FTL 500 may include a wear leveling module 510, a garbage collection module 520, an address mapping table 530, and a request generation module 540. The host FTL 500 may be used to at least partially control the OC-SSD. Therefore, the host FTL 500 may perform management operations on the OC-SSD, generate metadata associated with the management operations, and store the generated metadata in the OC-SSD. The wear leveling module 510 may manage the degree of wear of each storage medium included in the OC-SSD. The storage cells of the storage medium may age through write operations, erase operations, etc., and the aged storage cells (or worn storage cells) may cause defects. The wear leveling module 510 may manage erase cycles and programs corresponding to the storage cell array, thereby preventing a specific cell area from wearing out earlier than other cell areas. The garbage collection module 520 may arrange blocks storing data fragments. For example, in a flash memory device, an erase unit may be set to be larger than a programming unit, a programming operation and an erase operation may be repeated, and then an operation may be performed to collect valid data distributed in physically different blocks in the same block by using any free blocks. The free blocks may be generated based on the operation of the garbage collection module 520. In addition, in an OC-SSD environment, the host system may perform an address conversion operation between a logical address and a corresponding physical address. The mapping information between the resulting logical address and the physical address may be stored in the address mapping table 530. In addition, according to an embodiment, the request generation module 540 may generate various requests associated with the management of the meta area. For example, the request generation module 540 may generate a block allocation request including information about one or more blocks to be allocated to the meta area based on block information about the storage system. In addition, according to the embodiment described below, the host system may issue a request to additionally allocate a spare block for the meta area, and the request generation module 540 may generate an additional allocation request for additionally allocating a spare block. In addition, according to the above-described embodiment, the request generation module 540 may perform various processing operations associated with the meta area.
[0081] Figure 8 is a flow chart outlining an operating method for a data processing system according to an embodiment.
[0082] Reference Figure 8 , the data processing system may include a host system having a host FTL and an OC-SSD. According to an embodiment, the storage of metadata may be managed at least in part using the OC-SSD in a manner to ensure reliability of the metadata.
[0083] As an operation example, a storage system corresponding to an OC-SSD may be appropriately formatted and connected to a host system (S11). Such formatting and connection may be accomplished by performing an initialization operation on a data processing system. During the initialization process, information describing the storage system may be provided to the host system, such as information identifying the number and corresponding locations of high reliability blocks (EF) included in the storage system and information indicating repair operations supported by the storage system.
[0084] The host system may then issue a block allocation request Req_BLK to the storage system. In response, the storage system manages memory resources so that at least one high reliability block is designated for storing critical data such as metadata. As part of this management process, a repair operation may be applied to the metadata.
[0085] For example, according to the above embodiment, information identifying the number and / or location of high reliability blocks capable of storing metadata may be included in the block allocation request Req_BLK. Therefore, based on the block allocation request Req_BLK received from the host system, the storage system may allocate high reliability blocks from a meta region that may be used to store metadata (S12). According to the above embodiment, one or more spare blocks and one or more high reliability blocks for repairing errors in metadata may be allocated to the meta region.
[0086] Subsequently, when performing a storage operation, the host system may provide a write request for metadata to the storage system, and may transmit information indicating a block location where the metadata is to be written to the storage system (S13). The storage system may store the metadata in a block at a location indicated by the host system, and may perform a processing operation associated with repairing an error in the metadata (S14). For example, in the case of applying a data mirroring operation, the metadata written in a high reliability block indicated by the host system may be identically written to a corresponding spare block. Alternatively, in the case of applying a repair parity operation, repair parity information generated based on the metadata may be written to a spare block.
[0087] Subsequently, the host system may transmit a read request for reading metadata and information indicating a location of a block for storing metadata to the storage system ( S15 ).
[0088] The storage system may determine whether the block from which metadata is requested to be read corresponds to a bad block (S16). For example, the storage system may determine whether a defect occurs in the block periodically or during the process of writing and / or reading data, and based on the result of the determination, the storage system may determine the corresponding block as a bad block. In addition, when the block from which metadata is requested to be read corresponds to a bad block, the storage system may read metadata that is mirrored and stored in a spare block, or may repair the metadata by using a repair parity stored in the spare block and provide the repaired metadata to the host system (S17). On the other hand, when the corresponding block is not a bad block, metadata may be read from a high reliability block at a location requested by the host system (S18).
[0089] Fig. 9A and Fig. 9B is a related conceptual diagram showing a data processing system according to an embodiment in one example. Fig. 9A and Fig. 9B Together, an example of operation in which spare blocks are exhausted for ensuring the integrity of metadata is shown.
[0090] Reference Fig. 9A , the host system HOST may provide the OC-SSD with an access request for writing and reading metadata, and the OC-SSD may provide the read metadata to the host system HOST, or may transmit an access result such as information indicating that writing of the metadata has been completed to the host system HOST. In addition, the OC-SSD may perform a processing operation of ensuring the integrity of the metadata by using a spare block during access to the metadata, and may determine whether the spare block is exhausted.
[0091] Whether a spare block is exhausted can be determined based on various techniques. For example, when the meta area of the OC-SSD is in a state where it is difficult to ensure the integrity of the metadata, it can be determined that the spare block is exhausted. For example, in the case of applying data mirroring technology, the metadata of a high-reliability block (e.g., a first high-reliability block) can be stored in a corresponding spare block, and when the first high-reliability block is determined to be a bad block, the spare block corresponding thereto can be accessed. In this case, when the spare block is also determined to be a bad block, it may be difficult to ensure the integrity of the metadata stored in the first high-reliability block, in which case it can be determined that the spare block has been exhausted.
[0092] The host system HOST may receive the result of determining whether the spare blocks are exhausted, and may transmit an additional allocation request for additionally requesting spare blocks to the OC-SSD. According to an embodiment, the host system HOST may issue a request for additionally allocating spare blocks in consideration of a repair operation applied to the OC-SSD, and the number and location of spare blocks to be additionally allocated may be included in the additional allocation request. The OC-SSD may allocate additional blocks as spare blocks to the meta area, and may transmit the allocation result to the host system HOST.
[0093] Various situations may occur in the data processing system, for example, when an undesirable sudden power outage (SPO) occurs, the OC-SSD may store information associated with the above-mentioned high reliability blocks, spare blocks, and spare blocks allocated separately. Subsequently, when the data processing system is powered on, the OC-SSD may access metadata automatically or based on a request of the host system HOST, for example, since only blocks included in the meta area (e.g., high reliability blocks and spare blocks) are accessed, metadata may be read and provided to the host system HOST. For example, the host system HOST may access only blocks included in the meta area regardless of information from the OC-SSD, and thus metadata may be acquired. Alternatively, the OC-SSD may know information about blocks of the meta area, and therefore, the information may be provided to the host system HOST, thereby improving the efficiency of managing metadata.
[0094] In addition, refer to Fig. 9B , shows an example of allocating one or more spare blocks to the meta region. For example, when one or more high reliability blocks included in the meta region are determined to be bad blocks, and the spare blocks corresponding thereto are also determined to be bad blocks, additional spare blocks may be allocated to the meta region. Fig. 9B , it is shown that when the first high reliability block EF1 and the second high reliability block EF2 and the spare blocks Rsvd1 and Rsvd2 corresponding thereto are determined to be bad blocks, spare blocks Rsvd11 and Rsvd12 can be allocated based on the first high reliability block EF1 and the second high reliability block EF2.
[0095] Fig.10 is a block diagram illustrating a data processing system 600 according to an embodiment.
[0096] Reference Fig.10 , the data processing system 600 may include a host system 610 and a storage system 620, the host system 610 may include a host FTL 611, and the storage system 620 may include a controller 621 and a storage device 622. Fig.10Although not shown in the figure, the storage device 622 may include a plurality of storage media, and each storage medium may include a plurality of blocks. In addition, each storage medium may include at least one high reliability block. According to the above-described embodiment, the storage system 620 may transmit block information Info_BLK about blocks included in the storage medium to the host system 610, and the host system 610 may transmit a block allocation request Req_BLK including information about one or more blocks to be used for storing metadata to the storage system 620.
[0097] According to the above-described embodiment, management of blocks of the meta area may be performed by the storage system 620, and for this purpose, the controller 621 may include a bad block manager 621_1. The bad block manager 621_1 may determine whether a block associated with the storage metadata (or a block included in the meta area) among a plurality of blocks included in the storage device 622 is a bad block, and may perform management operations thereon. For example, when it is determined that a block of the meta area is a bad block, the bad block manager 621_1 may replace access to the bad block with access to another block, and may manage a block map including information associated with replacement of the block.
[0098] In addition, the host FTL 611 of the host system 610 may manage bad blocks in the blocks for storing user data, and therefore, the host FTL 611 may include a bad block manager associated with the user data. For example, determining whether a block is a bad block may be performed by the storage system 620, and the storage system 620 may provide the host system 610 with information about the block determined to be a bad block among the plurality of blocks storing the user data. The host system 610 may replace access to the block determined to be a bad block with access to another block, and the block mapping of the block storing the user data may be managed by the host system 610. In addition, the host system 610 may store metadata including the block mapping of the block storing the user data in a meta area of the storage device 622.
[0099] Fig.11A and Fig. 11B is a block diagram illustrating various examples of block allocation operations that may be performed in a storage system.
[0100] Some of the multiple blocks included in the storage device may be managed as high reliability blocks. According to the above embodiment, some of the multiple blocks of the storage device may correspond to high reliability blocks, for example, blocks storing data based on single-level cells may correspond to high reliability blocks, or blocks with good characteristics may be set as high reliability blocks based on test results autonomously performed by the storage system. Alternatively, in the process of manufacturing the storage system, processing performed on some blocks may be implemented to make their reliability relatively higher than that of other blocks, and the corresponding blocks may be set as high reliability blocks.
[0101] Reference Fig.11A , when setting the meta area Reg_Meta including the block to store metadata, when a high reliability block is included in the storage device, the reliability of data of the high reliability block may be relatively high and the possibility of error occurrence may be very low, and therefore, only the high reliability block may be included in the meta area Reg_Meta. In this case, the storage system or the host system may not allocate a spare block for a repair operation to the meta area Reg_Meta for storing metadata.
[0102] On the other hand, Fig. 11B As shown, in some embodiments, the storage device may include only a normal block BLK_N without managing a separate high reliability block, in which case a normal block BLK_N storing metadata and one or more spare blocks Rsvd for a repair operation corresponding to the metadata may be allocated to the meta region Reg_Meta. That is, although the probability of an error occurring in the normal block BLK_N is higher than that of the above-mentioned high reliability block, a repair operation may be applied through one or more spare blocks Rsvd, and therefore, the reliability of the metadata may be ensured.
[0103] Fig. 12A and Fig. 12B is a block diagram illustrating an example of allocating blocks to a meta area in an SSD including a plurality of storage media.
[0104] Reference Fig. 12A , the data processing system 700 may include a host system 710 and an SSD 720. The SSD 720 may correspond to the open channel SSD according to the above-described embodiment, transmit signals to and receive signals from the host system 710 through the signal connector SGL, and receive power through the power connector PWR. The SSD 720 may include an SSD controller 721, an auxiliary power supply 722, and a plurality of storage systems 723, 724, and 725 (i.e., Flash 1, Flash 2, and Flash n). The plurality of storage systems 723, 724, and 725 may each include one or more flash memory devices as storage devices. In addition, each flash memory device may include one or more bare die DIEs, and one or more blocks may be provided in each bare die DIE.
[0105] According to the above-described embodiment, the SSD controller 721 may communicate with the plurality of storage systems 723, 724, and 725 through the plurality of channels Ch1 to Chn, and may perform management operations on blocks storing metadata in the SSD 720. In addition, information associated with a block map may be generated through management operations performed on the blocks, and the information associated with the block map may be arbitrarily stored in the volatile memory 721_1 of the controller 721, and may be non-volatilely stored in the plurality of storage systems 723, 724, and 725. In addition, the controller 721 may include a bad block manager (not shown) for performing management operations on blocks storing metadata, and the bad block manager may be implemented based on hardware, software, or a combination thereof.
[0106] According to the above-mentioned embodiment, the meta area storing metadata may include high reliability blocks and spare blocks. For example, in the case where the repair operation uses data mirroring technology, a high reliability block (e.g., a first high reliability block) and a spare block corresponding thereto (e.g., a first spare block) may be selected, and the first high reliability block and the first spare block corresponding thereto may be selected based on various technologies. For example, the first high reliability block and the first spare block corresponding thereto may be selected in a storage system connected to the same channel, or may be selected in one of a plurality of flash memory devices included in one channel. Alternatively, each flash memory device may include a plurality of bare die DIEs, and the first high reliability block and the first spare block corresponding thereto may be selected in the same chip.
[0107] Fig. 12B An example of block selection according to an embodiment is shown, for example, a first flash memory device Flash 1 is shown as a flash memory device connected to a first channel Ch1. The first flash memory device Flash 1 may include a plurality of dies DIE, and a first die DIE1 723_11 and a second die DIE2 723_12 are shown among the plurality of dies. In addition, the first die DIE1 and the second die DIE2 may each include a plurality of blocks.
[0108] The host system may issue a request to allocate a plurality of high reliability blocks to the first flash memory device Flash 1, for example, a first high reliability block EF 1 set in the first die DIE 1 and a second high reliability block EF 2 set in the second die DIE 2 may be allocated to the meta area. In this case, one or more spare blocks to be used in connection with the repair operation may be further allocated to the meta area, for example, a first spare block Rsvd 1 corresponding to the first high reliability block EF 1 may be selected in the first die DIE 1, and a second spare block Rsvd 2 corresponding to the second high reliability block EF 2 may be selected in the second die DIE 2. For example, parallel communication may be performed between the controller 721 and the plurality of storage systems 723, 724, and 725, at which time spare blocks may be selected so that data access related to repairing errors in metadata does not affect data communication of another storage.
[0109] Fig.13 8 is a block diagram showing a network system 800 including a server system according to an embodiment. Fig.13 In the figure, a server system and a plurality of terminals (eg, computing nodes) are shown, and the server system can be implemented using the data processing system according to the above-described embodiments.
[0110] Reference Fig.13 , the network system 800 may include a server system 810 and a plurality of terminals 831_1 to 831_n that perform communication through a network 820. The server system 810 may include a server 811 and an SSD 812 corresponding to a storage system. The server 811 may perform the function of the host system according to the above-described embodiment.
[0111] The server 811 may process requests transmitted from a plurality of terminals 831_1 to 831_n connected to the network 820. For example, the server 811 may store data provided from the plurality of terminals 831_1 to 831_n in the SSD 812. In addition, when the data is stored in the SSD 812, metadata may be stored in the SSD 812 in order to manage the SSD 812. According to the above-described embodiment, the server 811 may issue a request to allocate a block for storing metadata to the SSD 812, and the SSD 812 may perform allocation on the block storing metadata, and may perform a bad block determination operation and a management operation on the allocated block. In addition, a repair operation for enhancing the reliability of the metadata may be applied.
[0112] While the inventive concept has been particularly shown and described with reference to embodiments thereof, it will be understood that various changes in form and details may be made therein without departing from the spirit and scope of the appended claims.
Claims
1. A storage system that communicates with a host system, wherein the host system is external to the storage system, and the storage system include: A storage device, the storage device comprising a storage medium, the storage medium being divided into a plurality of blocks including a high reliability block and a spare block; and A controller, the controller being configured to: sending, to the host system, block information identifying the high reliability block among the plurality of blocks and information indicating a plurality of repair operation techniques supported by the storage system, receiving a block allocation request from the host system, wherein the block allocation request identifies at least one of the high reliability blocks to be allocated to a meta-area storing metadata and includes a selected repair operation technique of the plurality of repair operation techniques selected by the host system, In response to the block allocation request, allocating the at least one high reliability block among the high reliability blocks to the meta region, setting the selected repair operation technique selected by the host system as a set repair operation technique for the high reliability block, and A different number of the spare blocks to be allocated to the meta region is selected based on the set repair operation technique, wherein: The controller includes: a bad block manager configured to manage a block allocation operation performed in response to the block allocation request; and a repair module configured to repair errors in the metadata stored in the at least one high reliability block in the high reliability blocks.
2. The storage system according to claim 1, in, The controller is further configured to: when one of the at least one of the spare blocks is exhausted, send information to the host system indicating that the one of the at least one of the spare blocks is exhausted, and The controller is further configured to, in response to a request from the host system requesting allocation of an additional spare block of the spare blocks, additionally allocate the additional spare block to the meta region.
3. The storage system according to claim 1, in, The selected repair operation technique includes a data mirroring technique that writes the metadata stored in the at least one of the high reliability blocks into the selected spare block.
4. The storage system according to claim 1, in, The selected repair operation includes a repair parity technique that writes repair parity information into the selected spare block, and The number of the at least one high reliability block among the high reliability blocks storing metadata is greater than the number of the spare blocks storing the repair parity information.
5. The storage system according to claim 1, in, The controller further includes an interface circuit configured to communicate with the host system using a non-volatile memory fast interface such that the block information and the block allocation request are transmitted via the interface circuit.
6. The storage system according to claim 1, in, The storage medium includes a plurality of flash memory devices managed by a flash translation layer running in the host system.
7. The storage system according to claim 1, in, The controller also includes a processor and a working memory accessible by the processor, and The bad block manager comprises a program loaded into the working memory and executed by the processor, wherein the program implements: a meta-area management module configured to allocate the at least one of the high-reliability blocks and at least one of the spare blocks to the meta-area in response to the block allocation request; and a bad block management module configured to manage block mapping information in response to a bad block determination performed on the at least one of the high reliability blocks of the meta region, The repair module is configured to perform a repair operation on the metadata stored in the at least one high reliability block among the high reliability blocks using information stored in the selected spare block.
8. The storage system according to claim 1, in, The storage system includes an open channel solid state drive configured to receive a physical address indicating a location at which the host system accesses the metadata.
9. The storage system according to claim 1, in, When a first repair operation technique is selected by the host system from among the plurality of repair operation techniques, a spare block is allocated corresponding to a high reliability block.
10. The storage system according to claim 9, in: When a second repair operation technique is selected by the host system from among the plurality of repair operation techniques, a spare block is allocated corresponding to at least two high reliability blocks, and The second repair operation technique has a lower error correction capability than the first repair operation technique.
11. A method for operating a storage system comprising a plurality of blocks, wherein the plurality of blocks comprises high reliability blocks and spare blocks, wherein the method include: sending block information identifying a number of the high reliability blocks in the plurality of blocks and information indicating a plurality of repair operation techniques supported by the storage system to a host system, the host system being external to the storage system; receiving a block allocation request from the host system, the block allocation request identifying a high reliability block among the high reliability blocks to be allocated to storage metadata based on the block information sent by the storage system and a spare block among the spare blocks corresponding to the high reliability block, the block allocation request further comprising a selected repair operation technique selected by the host system among the plurality of repair operation techniques; In response to the block allocation request, allocating the high reliability block to a meta region; In response to the block allocation request, allocating the spare block to the meta area; setting the selected repair operation technique selected by the host system as a set repair operation technique for the high reliability block; selecting a different number of the spare blocks to be allocated to the meta-region based on the set repair operation technique; writing metadata to the high reliability block in response to a metadata write request received from the host system; Determining whether the high reliability block is a bad block; and Once it is determined that the high reliability block is a bad block, the fact that the high reliability block is a bad block is communicated to the host system, and the metadata written in the high reliability block is repaired.
12. The operating method according to claim 11, further comprising: include: In response to the information indicating the data mirroring technique, writing the metadata to the selected spare block to mirror the metadata written to the high reliability block, wherein the metadata written to the selected spare block is used to repair the metadata stored in the high reliability block.
13. The operating method according to claim 11, further comprising: include: generating repair parity information from the metadata written in the high reliability block, and storing the repair parity information in the selected spare block in response to the information indicating a repair parity technique, The metadata written in the high reliability block is repaired using the repair parity information stored in the selected spare block.
14. The operating method according to claim 11, further comprising: include: determining whether the spare blocks have been exhausted; and Once it is determined that the spare block has been used up, information identifying the spare block as an exhausted spare block is transmitted to the host system.
15. A host system communicating with a storage system comprising a plurality of blocks, the host system being external to the storage system, the host system include: a host flash translation layer that manages the plurality of blocks, wherein the host flash translation layer includes a request generation module that generates a block allocation request for allocating at least one high-reliability block among the plurality of blocks to be used for storing metadata; and an interface circuit configured to be connected to the storage system through an interface to receive block information identifying the high reliability block and information indicating a plurality of repair operation technologies supported by the storage system during an initialization operation of the storage system, and to send a block allocation request from the host system to the storage system, wherein the block allocation request includes identifying the number of the high reliability blocks to be allocated to a meta area storing the metadata, a corresponding address location of each of the number of the high reliability blocks, and information indicating a selected repair operation technology selected by the host system from among the plurality of repair operation technologies, Wherein, different numbers of spare blocks from among the plurality of blocks to be allocated to the meta-area of the storage system are selected based on the selected repair operation technique.
16. The host system according to claim 15, in, The block allocation request also includes information identifying a number of spare blocks among the plurality of blocks to be used during a repair operation performed on the metadata stored in the number of the high reliability blocks.
17. The host system according to claim 15, in, The host system is configured to access the at least one of the high reliability blocks storing the metadata when the host system is powered on after being powered off.
Citation Information
Patent Citations
Rotary forge device, rotary forging method, method for manufacturing hub unit bearing using rotary forging method, and vehicle manufacturing method
KR1020190122654A
Three-Dimensional Semiconductor Memory Devices And Methods Of Fabricating The Same
US20110233648A1
Vertical-type non-volatile memory devices
US7679133B2
Non-volatile memory device, erasing method thereof, and memory system including the same
US8553466B2
Nonvolatile memory device, operating method thereof and memory system including the same
US8559235B2
