CN116601609A - store data in computer storage - Google Patents

Abstract

A method for storing data in a computer storage is disclosed. The method comprises the following steps: storing the data in a computer storage; storing erasure codes of the data in the computer storage; storing a copy of the data in the computer storage; based on the results of storing the copy of the data in the computer storage, the computer storage that acknowledges the erasure code storing the data is available to store other data. The erasure code of stored data may advantageously enable recovery of the data when it is corrupted. Disadvantageously, however, erasure codes occupy memory space. After storing the copy of the data, the memory storing the erasure code is identified as available for storing other data, advantageously freeing up memory space to store other data.

Description

store data in computer storage

technical field

This disclosure relates to storing data in computer storage.

Background technique

Data stored in computer storage systems can be protected from loss or damage by erasure coding or replication. Erasure coding creates a mathematical function that describes parts of the data as a function of other parts of the data, so that complete data can be recovered in the event of partial loss or corruption. Replication creates an additional copy of the data, which may be stored on a different storage device (eg, a different disk drive) than the original data. One advantage of erasure coding for protecting data is that the memory footprint of an erasure code is typically smaller than the footprint of the data it protects. One advantage of replication for protecting data is that the computational complexity of creating a copy and using that copy to restore the data is relatively low.

Contents of the invention

It is an object of the present disclosure to provide a method for storing data in computer storage, wherein said data is protected from loss or corruption.

The above and other objects are achieved by the features of the independent claims. Other implementations are evident from the dependent claims, the description and the figures.

A first aspect of the present disclosure provides a method for storing data in computer storage. The method comprises: storing data in computer storage; storing an erasure code of said data in said computer storage; storing a copy of said data in said computer storage; and storing said As a result of storing a copy of data in said computer storage, it is confirmed that said computer storage storing said erasure code for said data is available for storing other data.

In the methods described above, data stored in computer storage is protected by an erasure code (eg, parity information) for the data. Erasure coding of stored data can advantageously enable relatively fast protection of data. Therefore, the risk of data loss or corruption in the short term may be reduced. For example, the memory footprint of the erasure code is relatively small, and the erasure code can be stored in local storage, that is, stored in the local storage of the data. For example, when data is stored in one or more storage devices by the above method, even if one or more storage devices have relatively small spare storage capacity, erasure codes can also be stored in the same one or more storage devices . Thus, relatively quick protection of data can be achieved by avoiding the need to transfer data to an external backup resource and the delays associated with it. In addition, since erasure codes can be stored locally, in the event of data loss, data recovery using erasure codes may be relatively faster than recovering data from remote storage devices.

However, recovering data using erasure codes may be relatively computationally complex, and thus has the disadvantage of consuming computing resources of the storage system, eg, processor time. Therefore, the above method also includes: storing a copy of the data, that is, replicating the data. For example, a copy of data may be stored through a regular data backup process, whereby the copy data is stored on a different storage device than the original data. The creation and storage of replica data can be relatively long compared to erasure coding, for example, because the memory footprint of replica data is relatively large and the replica data needs to be transferred to a suitably large storage device that can be kept away from the data storage location. However, replicating the data can advantageously make recovering the data computationally simple in the event that stored data is partially lost or corrupted.

Thus, erasure coding of stored data combined with subsequent copies of stored data may advantageously provide early protection of data by erasure coding and enhanced long-term protection of data by replicating data.

However, after a copy of the data is stored, that is, after the data has been replicated, erasure coding can be considered redundant, since the copy of the data can then be used to recover the data in the event of data loss or corruption. At the same time, erasure codes take up storage space, which can be particularly problematic when erasure codes are stored in storage devices with relatively small storage capacities. Accordingly, the method described above includes the further step of confirming that the computer storage storing the erasure code is available for storing other data as a result of storing a copy of the data in the computer storage. In other words, in the above method, once the data is copied and the copy data is stored, the above method includes: confirming that the storage space occupied by the erasure code can be covered by new data. In an example, the above method may even include: erasing the erasure code from the storage after storing the copy data. Thus, by this step of the method described above, the total memory footprint of data and protection configurations can advantageously be reduced. However, since this step of the above method is performed as a result of storing a copy of the data, the above method is safe because the execution of this step depends on the storage of the copy data. Therefore, the risk of data having neither erasure codes nor duplicate data is reduced.

In an example, the method may include storing two or more of the data, the erasure code, and the duplicate data via a common storage device, eg, at different locations on the same disk drive. In other examples, two or more of the data, the erasure code, and the duplicate data may be stored by mutually different storage devices. For example, data and erasure codes may be stored on a first set of one or more storage devices (e.g., disk drives), while duplicate data may be stored on one or more different storage devices (e.g., disk drives) .

The above method may include the step of creating an erasure code, eg, parity information, before storing the erasure code. Methods for creating erasure codes for data are known to those skilled in the art.

In the context of this specification, unless otherwise stated, the terms "storage" (eg, computer memory) and "internal memory" (eg, computer memory) are used to refer generally to storage resources used to store data. In this regard, each term is intended to include storage resources used for long-term storage of data (eg, magnetic disk drives, solid-state drives, or flash memory) as well as storage resources used for short-term storage of data (eg, random access memory).

In one implementation, said storing data in computer storage includes: storing a first segment of said data in a first computer storage device, and storing a second segment of said data in a second computer storage device. in the device.

In other words, the above method may include: distributing data among multiple storage devices, which is called striping hereinafter. Striping of data can advantageously improve data read/write performance because segments of data can be simultaneously read from/written to their respective storage devices. Additionally, striping data across multiple devices reduces the proportion of data that would be lost if a subset of storage devices failed. Therefore, erasure codes are only resilient to loss of relatively small amounts of data, which is usually acceptable. This can advantageously reduce the complexity of erasure codes, thereby reducing the complexity of generating and using erasure codes to recover data. In an example, the above method may include storing data in two or more segments, optionally in two or more storage devices. For example, in an example, the above method may include: storing data in five or six segments in five or six storage devices. The above method may include the following steps: before storing the data, divide the data into two or more segments.

In one implementation, the storing a copy of the data in the computer storage includes: storing a copy of the data in a separate computer storage device from which the data is stored and/or from a storage device storing the corrected data. Erasure coded computer storage devices separate one or more computer storage devices.

In other words, in an example, the above method includes: storing the duplicate data in a storage device physically separate from at least one of the storage device storing the data and the storage device storing the erasure code. This separation of replica data from data and/or erasure codes advantageously reduces the likelihood of losing replica data and data or erasure codes in the event of a storage device failure. In an example, data and erasure codes may be stored on one or more storage devices (eg, hard drives), while duplicate data may be stored on one or more separate storage devices.

In one implementation, the method further includes storing location data identifying a location of the copy of the data in the computer storage.

In other words, the location data defines the machine-readable address of the replica data in computer storage. For example, the location data may identify the storage device number and/or offset where the duplicate data resides. Thus, location data may advantageously enable easy retrieval of duplicate data from computer storage, which may be useful in the event of data loss. For example, location data can be generated by the memory controller when storing replica data.

In one implementation, said storing location data identifying a location of said copy of said data in said computer storage comprises: storing said location data in said computer storage storing said copy of said data in the device.

In other words, the location data may be stored on the same storage device as at least a portion of the replica data. The advantage of this setup is that the location data can be easily accessed by the memory controller used to access the replica data. Therefore, the same memory controller can read the location data and retrieve the duplicate data when a data loss causes the duplicate data to be needed. This can advantageously reduce the time required to restore data using duplicate data. For example, duplicate data may be stored on a single storage device, while location data may be stored on that same storage device. In this example, the location data may identify an offset within the storage device of the duplicate data. In another example, replica data may be distributed across multiple storage devices, and location data may be stored on one of the multiple devices.

In one implementation, the method further includes storing in the computer storage a flag marking the computer storage storing the data as a result of the storing a copy of the data in the computer storage in storage.

In other words, the above method may include: after storing the replica data, marking a location in the storage where the data is stored. Thus, a flag can identify a location of computer storage, including all or part of data, which has been protected by replication (i.e. by storing duplicate data). Therefore, the flag can be used to easily confirm whether the data has been copied, that is, whether the duplicate data has been stored. One advantage of this is that the memory manager can simply read the flag to determine whether the data has been copied, so in the event of data loss it knows whether the copied data can be retrieved.

In one implementation, the method further includes storing other data in the computer storage identified as being available for storing other data. In other words, the above method may further include: after copying the data, overwriting one or more locations in the computer storage where the erasure code is stored.

In one implementation, the storing the erasure code of the data in the computer storage includes: storing the erasure code in one or more computer storage devices storing the data. In other words, erasure codes can be stored on the same storage device or devices as the data. Given that erasure codes are generated from data, this setup reduces the time required to generate and store erasure codes, since the erasure code generator can read from and write to the same storage device or devices. one or more storage devices.

In an implementation manner, the storing the erasure code of the data in the computer storage includes: storing the first erasure code and the second erasure code in the computer storage, wherein the The first erasure code and the second erasure code are independent of each other; and according to the result of said storing a copy of said data in said computer storage, validating said erasure code storing said data Said computer storage being available to store other data comprises: as a result of said storing a copy of said data in said computer storage, confirming that only said computer storage storing said second erasure code of said data is available to store other data.

In other words, the above method may include: storing two mutually independent erasure codes, that is, each erasure code can operate independently of other erasure codes on each part of the data, so as to recreate the data. Creating/storing two erasure codes for data can advantageously improve data resilience compared to a single erasure code, since one of the erasure codes itself may be lost or corrupted, while the data can still be recovered using the remaining erasure code recover. However, in the implementation of the above method, only one of the erasure codes (that is, the second erasure code) is selected to be overwritten after the data is copied, and the second erasure code is kept in storage. This setup provides a favorable balance of data resiliency versus protected memory footprint. In other examples where multiple erasure codes are generated, one may choose to overwrite all erasure codes after copying the data.

In one implementation, the method further includes: performing an error detection operation on the data stored in the computer storage; in response to detecting an error in the data stored in the computer storage, retrieving the stored said copy of said data in said computer storage.

In other words, in an example, the above method may include testing the data to check for errors, eg, partial or total loss or corruption of the data, and in response to detecting data loss/corruption, the above method may include retrieving duplicate data. This process of "proactively" testing data and retrieving replica data can reduce the time between data loss and using the replica data to restore the data, thereby advantageously minimizing the time for clients to access the data after data loss.

A second aspect of the disclosure provides a computer program. The computer program includes instructions that, when executed by a processor, cause the processor to: store data in computer storage; store an erasure code for the data in the computer storage; storing a copy of said data in said computer storage; confirming that said computer storage storing said erasure code of said data is available as a result of said storing said copy of said data in said computer storage to store other data.

In an implementation manner of the second aspect, the computer program includes instructions, and when the instructions are executed by a processor, the processor executes the method provided by any implementation manner of the first aspect of the present disclosure. .

A third aspect of the disclosure provides a computer readable data carrier. The computer program provided by the second aspect of the present disclosure is stored in the computer-readable data carrier.

A fourth aspect of the present disclosure provides a computing system. The computing system includes computer storage and a processor coupled to the computer storage, the processor configured to: store data in the computer storage; store an erasure code for the data in the computer storage; a copy of said data is stored in said computer storage; and based on the result of said storing a copy of said data in said computer storage, confirming that said computer storage storing said erasure code for said data is available for storage other data.

In an implementation manner of the third aspect, the processor is configured to execute the method provided in any implementation manner of the first aspect of the present disclosure.

These and other aspects of the disclosure are apparent in one or more of the embodiments described below.

Description of drawings

In order to understand the present disclosure more easily, embodiments of the present disclosure are described below by way of examples in conjunction with the accompanying drawings, in which:

FIG. 1 is a schematic diagram of one example of a computing system embodying an aspect of the present disclosure;

2 is a schematic diagram of computer memory in a computing system;

3 is a schematic diagram of a computer storage device in a computing system;

4 is a schematic diagram of processes involved in an exemplary method for storing data in a computer storage device identified with reference to FIG. 3, the method including a method of partitioning data for storage;

FIG. 5 is a schematic diagram of a process involved in a method of segmenting data for storage;

6 is a schematic diagram of a process of storing location data;

Figure 7 is a schematic diagram of the process of storing the flags in computer storage;

Figure 8 is a schematic diagram of a process for confirming that computer storage storing erasure codes can be used to store other data;

Figure 9 is a schematic illustration of the process of storing other data at identified locations in computer storage; and

FIG. 10 is a schematic diagram of a process of performing an error detection operation.

Detailed ways

Referring initially to FIGS. 1 and 2 collectively, an example computing system 101 embodying an aspect of the present disclosure includes a plurality of client devices 102 and 103 , a computer storage system 104 and a backup storage system 105 . Client devices 102 and 103 are communicatively coupled to computer storage system 104 through network 106 . Computer storage system 104 is communicatively coupled to backup storage system 105 through network 107 .

Client devices 102 and 103 use computer storage system 104 to store data. For example, client devices 102 and 103 may send requests to computer storage system 104 over network 106 to store or retrieve data, and may similarly exchange data with computer storage system 104 over network 106 for storage or retrieval. Client devices 102 and 103 may be, for example, desktop computers, laptop computers, or smartphones. Client devices 102 and 103 may include computing functionality, such as a computer processor. In the above example, two client devices 102 and 103 are depicted as being coupled to computer storage system 104 . In other examples, the number of client devices coupled to computer storage system 104 may be greater or less than two. Client devices 102 and 103 may be remote from computer storage system 104, and indeed relatively remote from each other. For example, computer storage system 104 may be located in a central data center. Accordingly, client devices 102 and 103 may use storage resources in computer storage system 104 to store data.

Computer storage system 104 is used to provide storage resources for multiple client devices (eg, client devices 102 and 103). In the above example, computer storage system 104 is used to communicate with client devices 102 and 103 for storing and retrieving data, and also to communicate with backup storage system 105 for storing data in computer storage system 104, as described below. A copy of , that is, stored in the backup storage system 104.

Computer storage system 104 includes processor 108 , memory 109 , storage 110 , input/output interface 111 , and system bus 112 . Processor 108 is used to control the operation of the computer storage system, eg, to process storage and retrieval requests from client devices 102 and 103 . As described herein, in an example, processor 108 is configured to control the course of data storage operations according to a data storage computer program stored in memory 109 . The memory 109 is configured as a nonvolatile read/write memory for storing computer programs executed by the processor 108 and operation data associated with operations performed by the processor 108 . In an example, the memory 109 stores a computer program 201 for controlling data storage. In an example, memory 109 is flash memory, but in other examples, flash memory may be replaced by an alternative form of memory. The memory 110 is used for storing data, for example, for storing data of the client devices 102 and 103 . In an example, as described in detail below with reference to subsequent figures, memory 110 includes a plurality of storage devices, which in an example are a plurality of disk drives. Input/output interface 111 is used to connect client devices 102 and 103 to computer storage system 104 , and to connect computer storage system 104 to backup storage and deduplication system 105 . Components 108 through 111 in computer 104 communicate over system bus 112 .

The backup storage system 105 is used to provide backup storage resources to the computer storage system 104 , that is, to store one or more copies of data stored in the computer storage system 104 . Backup storage of data provides useful redundancy, allowing data to be recovered if the original data is lost or corrupted (eg, in the event of memory 110 failure). The backup storage system 105 includes a processor 112 , a memory 113 , an input/output interface 115 and a system bus 116 . The processor 112 is used to control the operation of the backup storage system 105 . As described herein, in an example, processor 112 is used to control the process of communicating with computer storage system 104 and to cooperate with computer storage system 104 to control data backup operations. Memory 113 is used for non-volatile storage of data, eg, for storing copies of data received from computer storage system 104 for backup. In an example, storage 113 includes one or more disk drives. Input/output interface 115 is used to connect backup storage system 105 to computer storage system 104 . Components 112 to 115 communicate via system bus 116 .

Backup storage system 105 may be remote from computer storage system 104 . In practice, to provide backup storage for computer storage system 104, it may be desirable for backup storage system 105 to include physical storage resources, such as one or more disk drives, separate from the storage resources in computer storage system 104 to avoid Risk of simultaneous failure of two storage resources. Specifically, in some examples it may be advantageous for backup storage system 105 to be geographically remote from storage system 104, thereby reducing the risk of simultaneous failure of both storage systems.

In an example, both the networks 106 and 107 can be composed of a wide area network (wide area network, WAN) such as the Internet, a local area network (local area network, LAN), a metropolitan area network (metropolitan area network, MAN) and/or a personal area network (personal area network). , PAN) etc. to achieve. The above two networks may use wired technologies (e.g., Ethernet, Data Over Cable Service Interface Specification (DOCSIS), Synchronous Optical Networking (SONET) and/or Synchronous Digital Hierarchy , SOH), etc.) and/or wireless technologies (e.g., Institute of Electrical and Electronics (IEEE) 802.11 (Wi-Fi), IEEE 802.15 (WiMAX), Bluetooth, ZigBee, Near Field Communication ( near-field communication, NFC) and/or long-term evolution (Long-Term Evolution, LTE), etc.). The above two networks may include at least one device for data communication in the network. For example, both networks 106 and 107 may include computing devices, routers, switches, gateways, access points, and/or modems.

Referring next to FIG. 3 , in an example, memory 110 in computer storage system 104 includes a plurality of storage devices for storing data. In the above example, memory 110 includes five separate storage devices 301 to 305, which in the example are disk drives. Five memory devices 301 through 305 are coupled to system bus 116 through system bus 306 to act as logical units that are individually addressable by processor 108 . In an example, as described in detail below with reference to subsequent figures, computer program 201 executed by processor 108 causes data to be stored in memory 110 in multiple segments, with each segment in the data structure being stored on multiple disk drives. on a different disk drive. In an example, the data segments are stored in a redundant array of independent disks (RAID) level 6 configuration, e.g., by block-level striping of the data segments across multiple storage devices, as described below, where , two check blocks are distributed on the storage device.

Thus, in the above example, data items (eg, files) A to E are all stored in memory 110 . Data items A to E are each divided into three segments or blocks 1 to 3, which may be of equal size. For each data item A to E, the constituent segments or blocks 1 to 3 are striped, i.e. distributed, over respective groups of three storage devices such that each storage device holds at most three segments/blocks of data items A to E respectively one of. In addition, in an example, for each data item A to E, two erasure codes (eg, parities Pp and Pq) are stored in the memory 110 . In the example, for each data item A to E, data segments/blocks 1 to 3 and respective parities Pp and Pq are distributed across storage devices 301 to 305 such that each storage device holds at most these data segments/blocks or one of checksums.

As detailed below, the parities Pp and Pq of each data item A to E are configured independently of each other such that both parities Pp and Pq can be independent of other parities on their respective subsets of data segments/blocks 1 to 3. Perform operations to restore data. An advantage of this configuration is that the memory 110 has a relatively strong recovery capability for a subset of lost segments/blocks or corresponding parities. Specifically, in the configuration described above, for each data item A to E, the memory is resilient to the loss of at most two of the segments/blocks 1 to 3 or parities Pp and Pq, since data can usually be retrieved from Any three of block/segment or parity to rebuild.

Referring next to FIG. 4 , in an example, computer program 201 stored in memory 109 of computer storage system 104 is executed by computer storage system 104 to cause processor 108 to perform a method comprising four stages.

In stage 401 , the computer program 201 may cause the processor 108 to store data in the memory 110 of the storage system 103 . For example, stage 401 may be initiated in response to a storage request received by the storage system from one of client devices 102 and 103 , ie a request by the corresponding client device to store data in storage system 104 . In response to this request, processor 108 may then store the received data in memory 110 . In an example, as described above with reference to FIG. 3, and as shown in detail in FIG. These segments are striped on the storage devices 301 to 305 of . Thus, for example, stage 401 may comprise that the processor 108 divides each of the plurality of files of received data (e.g., files A to E) into a plurality of blocks/segments, e.g., three blocks/segments 1 to 3, and The segments are striped across three of the storage devices 301-305. Therefore, in an example, stage 401 may include three stages 501 to 503 to strip data, as shown in FIG. 5 .

In stage 402 , computer program 201 may cause processor 108 to create an erasure code, eg a parity check, for the data stored in stage 401 and also store the erasure code in memory 110 . In an example, stage 402 may include the processor 108 creating two mutually independent erasure codes for the data. Suitable procedures for creating erasure codes (eg, parity) for data are known to those skilled in the art. For example, for each data item (e.g., file) A to E, the parity "p" and "q" can be calculated using two different functions, for example, the parity "p" can be calculated using the XOR function according to the array data, the check "q" can be calculated using a Reed-Solomon (Reed-Solomon) code. Therefore, erasure codes (eg, parity checks) can be resilient to failures of at most two of the storage devices 301-305.

In stage 403 , computer program 201 may cause processor 108 to perform a copying process by creating and storing in memory 113 of backup storage system 105 a copy of the data stored in stage 401 . In an example, there may be a delay between performing stage 402 and stage 403 . In an example, stage 401 and stage 402 in the above method may be executed repeatedly, for example, in response to repeated storage requests from client devices 102, 103, while stage 403 may only be executed after some instances of stage 401 and stage 402 are executed. For example, the backup operation of stage 403 may be performed periodically, eg, daily, such that copies of all data stored in storage system 104 during this period may be backed up within a single time step. This could be a computationally efficient method of data replication. Thus, in stage 403, processor 108 in storage system 104 may create a copy of one or more data items stored in memory 110 in one or more stages 401, and may send the copy data over network 107 to Backup storage system 105 . The processor 108 in the storage system 104 can communicate with the processor 112 in the backup storage system 105 so that the backup storage system 105 saves the copy data in the memory 113 . Stage 403 may include: the backup storage system 105 sends a report back to the storage system 104 to confirm the storage of the duplicate data.

In an example, stage 403 may also include the processor 108 causing location data identifying the location of the replica data in the memory 113 of the backup storage system 105 to be stored. For example, referring to FIG. 6 , in an example, stage 403 may include the process of step 601 , ie storing the location data in memory 113 together with the copy data. Accordingly, the location of the duplicate data in memory 113 may be conveniently identified, for example, by processor 108 or 112 when the duplicate data needs to be retrieved from memory.

In an example, stage 403 may also include storing in memory 110 of storage system 104 one or more flags marking locations in computer memory 110 where data copied in stage 403 is stored. Thus, these flags may identify, for example, to processor 108 , portions of data stored in memory 110 that have been replicated, ie data for which a backup copy has been stored in backup storage system 105 . Therefore, when the data in the memory 110 is lost, the processor 108 can confirm whether there is a copy of the data in the backup storage system 105, so as to quickly retrieve the copy data.

In stage 404, the computer program 201 may cause the processor 108 to confirm that the location in the memory 110 for storing the erasure code for the corresponding data in stage 402 is available as a result of storing a copy of the data in backup storage in stage 403 to store other information. In other words, the data is replicated in stage 403 so that a copy of the data is stored in the backup system 105, after which the processor 108 in the storage system 104 can determine that the erasure coding of the data stored in the memory 110 is no longer needed , and the location occupied by the erasure code in memory can be overwritten by other (eg, new) data. For example, processor 108 may reserve registers in memory 110 that are locations in memory 110 that are available (ie, may allow use) to store other data. In an example, this stage may thus include the processor 108 modifying the register to confirm in the register that the location where the erasure code was stored in stage 402 can be used to store other data. Thus, at a later time step, when processor 108 wishes to ascertain an available location in memory 110 for storing other data, the processor may consult the register. In other examples, stage 404 may include: the processor 108 sets a flag on the location where the erasure code is stored in the memory 110, and the flag is used to identify a storage location that can be used to store other data. In an example, stage 404 may include processor 108 causing erasure codes stored in stage 402 to be erased from their corresponding locations in memory 110 .

In an example, stage 404 may comprise the process of stage 801 described with reference to FIG. 8 . Thus, stage 404 may include erasing only one set of erasure codes (e.g., parity "p"), while even after copying the data in stage 403, erasing a second set of erasure codes (e.g., parity "p") q") is stored in the memory 110. This may advantageously balance the conflicting needs of minimizing the memory footprint and maintaining the resiliency of data stored in memory 110 .

In an example, stage 404 may comprise the process of stage 901 described with reference to FIG. 9 . Accordingly, stage 404 may comprise a further step of storing other data in the computer memory 110 at a location identified in stage 404 as available for storing the other data.

Referring finally to FIG. 10 , in an example, computer program 201 run by processor 108 may cause processor 108 to perform error detection operation 1001 whereby processor 108 checks the integrity of data stored in memory 110 to check whether the data lost or damaged. When the processor detects data loss or corruption, the processor 108 may then retrieve a copy of the lost/corrupted data from the memory 113 of the backup storage 105 . In other words, in an example, the above method may include testing the data to check for errors, eg, partial or total loss or corruption of the data, and in response to detecting data loss/corruption, the above method may include retrieving duplicate data. This process of "proactively" testing data and retrieving replica data can reduce the time between data loss and using the replica data to restore the data, thereby advantageously minimizing the time for clients to access the data after data loss.

Various aspects of the present disclosure have been described herein in the context of a computing system that includes a primary storage system 104 serving remote client devices 102 and 103 and a backup storage system 105 remote from and serving storage system 104 . However, aspects of the present disclosure have broader applicability than this configuration. For example, in a simple example of various aspects of the present disclosure, computing system 101 may comprise a monolithic device, and data, one or more erasure codes, and replicated data may all be stored in the monolithic device's storage. In an example, the data, one or more erasure codes, and replica data may all be stored on a single storage device, eg, on a single disk drive. Such a computing system may include computer storage, e.g., one or more storage devices, e.g., disk drives, and a processor coupled to the storage for storing data in the computer storage, erasure coding the data storing in computer storage, storing a copy of data in computer storage, and based on the results of storing a copy of data in computer storage, confirming that erasure coded computer storage storing data can be used to store other data.

Although various aspects of the present disclosure and their associated advantages have been described in detail, it should be understood that various changes, substitutions and substitutions can be made herein without departing from the spirit and scope of the invention as defined by the appended claims. and change. In the claims, the word "comprising" does not exclude other elements or steps, and "a" or "an" does not exclude a plurality. Although the procedures of the methods provided by this disclosure have been described herein as occurring in a particular order, in other examples of the disclosure, the procedures of these methods may be performed in an alternate order, or may even be omitted from the methods.

Claims (15)

1. A method for storing data in a computer storage, the method comprising:

storing the data in a computer storage;

storing erasure codes of the data in the computer storage;

storing a copy of the data in the computer storage; and

based on the results of storing the copy of the data in the computer storage, the computer storage that acknowledges the erasure code storing the data is available to store other data.

2. The method of claim 1, wherein storing the data in a computer store comprises: a first segment of the data is stored in a first computer storage device and a second segment of the data is stored in a second computer storage device.

3. The method of claim 1 or 2, wherein the storing the copy of the data in the computer storage comprises: the copy of the data is stored in one or more computer storage devices separate from the computer storage device storing the data and/or separate from the computer storage device storing the erasure code.

4. The method according to any of the preceding claims, wherein the method further comprises: location data identifying the location of the copy of the data is stored in the computer storage.

5. The method of claim 4, wherein the storing location data in the computer storage that identifies a location of the copy of the data comprises: the location data is stored in a computer storage device that stores the copy of the data.

6. The method according to any of the preceding claims, wherein the method further comprises: and storing a flag of the computer storage that marks the storage of the data in the computer storage according to the result of storing the copy of the data in the computer storage.

7. The method according to any of the preceding claims, wherein the method further comprises: other data is stored in the computer storage that confirms that other data is available for storage.

8. The method of any of the preceding claims, wherein the storing the erasure code of the data in the computer storage comprises: the erasure code is stored in one or more computer storage devices that store the data.

9. The method of any of the preceding claims, wherein the storing the erasure code of the data in the computer storage comprises: storing a first erasure code and a second erasure code in the computer storage, wherein the first erasure code and the second erasure code are independent of each other; based on the results of storing the copy of the data in the computer storage, validating the computer storage storing the erasure code of the data is available for storing other data including: based on the results of storing a copy of the data in the computer storage, it is confirmed that only the computer storage storing the second erasure code of the data is available for storing other data.

10. The method according to any of the preceding claims, wherein the method further comprises: performing an error detection operation on the data stored in the computer storage; in response to detecting that the data stored in the computer storage is in error, the copy of the data stored in the computer storage is retrieved.

11. A computer program comprising instructions which, when executed by a processor, cause the processor to:

storing the data in a computer storage;

storing erasure codes of the data in the computer storage;

storing a copy of the data in the computer storage; and

based on the results of storing the copy of the data in the computer storage, the computer storage that acknowledges the erasure code storing the data is available to store other data.

12. The computer program according to claim 11, characterized in that the computer program comprises instructions which, when executed by a processor, cause the processor to perform the method according to any of claims 2 to 10.

13. A computer readable data carrier, characterized in that the computer readable data carrier has stored therein a computer program according to claim 11 or 12.

14. A computing system comprising a computer storage and a processor coupled to the computer storage, the processor configured to:

storing data in the computer storage;

storing erasure codes of the data in the computer storage;

storing a copy of the data in the computer storage; and

based on the results of storing the copy of the data in the computer storage, the computer storage that acknowledges the erasure code storing the data is available to store other data.

15. The computing system of claim 14, wherein the processor is configured to perform the method of any of claims 2-10.