Micro-service based deduplication

Abstract

A system that manages an object storage may include frontend micro-services and backend micro-services. The frontend micro-services may obtain a request to store data in an object storage and divide the data into slices. The backend micro-services may generate a sketch of each slice, match each slice to a similarity group of a plurality of similarity groups, obtain meta-data associated with each matched similarity group, and add at least a portion of a slice of the slices to a compression region using the meta-data.

Description

BACKGROUND

Computing devices generate, use, and store data. The data may be, for example, images, document, webpages, or meta-data associated with any of the files. The data may be stored locally on a persistent storage of a computing device and/or may be stored remotely on a persistent storage of another computing device.

SUMMARY

In one aspect, a system for managing an object storage in accordance with one or more embodiments of the invention may include frontend micro-services and backend micro-services. The frontend micro-services may obtain a request to store data in an object storage and divide the data into slices. The backend micro-services may generate a sketch of each slice, match each slice to a similarity group, obtain meta-data associated with each matched similarity group, and add at least a portion of a slice of the slices to a compression region using the meta-data.

In one aspect, a method of managing an object storage in accordance with one or more embodiments of the invention may include obtaining, by a frontend micro-service, a request to store data in an object storage; dividing, by the frontend micro-service, the data into slices; generating, by a backend micro-service, a sketch of each slice; matching, by the backend micro-service, each slice to a similarity group; obtaining, by the backend micro-service, meta-data associated with each matched similarity group; and adding, by the backend micro-service, at least a portion of a slice of the slices to a compression region using the meta-data.

In one aspect, a non-transitory computer readable medium in accordance with one or more embodiments of the invention includes computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for managing an object storage, the method includes obtaining, by a frontend micro-service, a request to store data in an object storage; dividing, by the frontend micro-service, the data into slices; generating, by a backend micro-service, a sketch of each slice; matching, by the backend micro-service, each slice to a similarity group; obtaining, by the backend micro-service, meta-data associated with each matched similarity group; and adding, by the backend micro-service, at least a portion of a slice of the slices to a compression region using the meta-data.

BRIEF DESCRIPTION OF DRAWINGS

Certain embodiments of the invention will be described with reference to the accompanying drawings. However, the accompanying drawings illustrate only certain aspects or implementations of the invention by way of example and are not meant to limit the scope of the claims.

FIG. 1 shows a diagram of a system in accordance with one or more embodiments of the invention.

FIG. 2A shows a diagram of an object storage in accordance with one or more embodiments of the invention.

FIG. 2B shows a diagram of a similarity group meta-data entry in accordance with one or more embodiments of the invention.

FIG. 2C shows a diagram of an object recipe storage in accordance with one or more embodiments of the invention.

FIG. 2D shows a diagram of a slice recipe storage in accordance with one or more embodiments of the invention.

FIG. 3A shows a diagram of data in accordance with one or more embodiments of the invention.

FIG. 3B shows a diagram of slices of the data of FIG. 3A in accordance with one or more embodiments of the invention.

FIG. 3C shows a diagram of segments of a slice of FIG. 3B in accordance with one or more embodiments of the invention.

FIG. 4 shows a flowchart of a method of storing data in an object storage in accordance with one or more embodiments of the invention.

FIG. 5 shows a flowchart of a second method of storing data in an object storage in accordance with one or more embodiments of the invention.

FIG. 6 shows a flowchart of a method of reading data stored in an object storage in accordance with one or more embodiments of the invention.

FIG. 7 shows a flowchart of a second method of reading data stored in an object storage in accordance with one or more embodiments of the invention.

FIG. 8 shows a flowchart of a method of adjusting the quantity of computing resources assigned to frontend micro-services or backend micro-services in accordance with one or more embodiments of the invention.

DETAILED DESCRIPTION

Specific embodiments will now be described with reference to the accompanying figures. In the following description, numerous details are set forth as examples of the invention. It will be understood by those skilled in the art that one or more embodiments of the present invention may be practiced without these specific details and that numerous variations or modifications may be possible without departing from the scope of the invention. Certain details known to those of ordinary skill in the art are omitted to avoid obscuring the description.

In general, embodiments of the invention relate to methods and systems for managing an object storage. More specifically, the methods and systems may provide functionality for deduplicating data before storing the data in the object storage. Deduplicating the data, before storing the data in the object storage, may increase the amount of data that can be stored in the object storage when compared to the amount of data that can be stored in the object storage without deduplicating the data. Deduplicating the data may also decrease the cost associated with storing data in the object storage by reducing the total amount of storage required to store the deduplicated data when compared to the amount of storage required to store the data without being deduplicated.

As used herein, deduplication refers to methods of storing only portions of data that are not already stored in the storage. For example, when multiple versions of a large text document, having only minimal differences between each of the versions, are stored without deduplication, storing each version will require approximately the same amount of storage space of a persistent storage. In contrast, when the multiple versions of the large text document are stored with deduplication, only the first version of the multiple versions stored will require a substantial amount of storage. Once the first version is stored in a persistent storage, the versions of the large word document subsequently stored will be deduplicated before being stored in the persistent storage resulting in much less storage space of the persistent storage being required to store the subsequently stored versions when compared to the amount of storage space of the persistent storage required to store the first stored version.

In one or more embodiments of the invention, the method of deduplication may include receiving data and dividing the data into slices by one or more frontend micro-services. The slices may be deduplicated by one or more backend micro-services by matching each slice to a similarity group. The meta-data associated with the matched similarity group may specify fingerprints of a subset of segments of all the segments stored in the object storage. For each slice, a fingerprint of each segment of the slice may be matched to the fingerprints of the subset of the segments. Any fingerprints that are not matched to any of the fingerprints of the subset of the segments may be added to one or more compression regions. Fingerprints, corresponding to each fingerprint that was not match, maybe added to the meta-data of the similarity group. The one or more compression regions may be stored in the object storage. Slice recipes and an object recipe that allow for each slice of the data and the data, respectively, to be reconstructed from the data stored in the object storage may be stored for future use.

In one or more embodiments of the invention, the number of frontend micro-services and/or the number of backend micro-services may be dynamically adjustable. Adjusting the number of each micro-service may enable computing resources to be efficiently allocated to different portions of the method of deduplication. More efficiently allocating the computing resources used to store data in the storage may improve the data storage throughput rate of the storage, eliminating bottlenecks of the storage, decrease the cost of storage by dynamically deallocating reserved computing resources in response to changes in rates of storage, and/or provide a scalable data storage system.

FIG. 1 shows a system in accordance with one or more embodiments of the invention. The system may include clients (100) that store data in an object storage (160) of a persistent storage (150). The system may include a deduplication service (110) that deduplicates the data from the clients (100) before the data is stored in the object storage (160). In one or more embodiments of the invention, the clients (100) and/or the persistent storage (150) are not part of the system.

The clients (100) may be computing devices. The computing devices may be, for example, mobile phones, tablet computers, laptop computers, desktop computers, or servers. Further examples of clients (100) include clients that are running in the cloud, such as on a virtual machine or in a container running in a public and/or private cloud environment. The clients (100) may be other types of computing devices without departing from the invention.

The clients (100) may be operably connected to the deduplication service (110). While not shown in FIG. 1, the clients (100) may be operably connected to the persistent storage (150) or other elements of the system without departing from the invention.

The persistent storage (150) may be hard disk drives, solid state drives, any other type of persistent storage media, or a logical storage including any combination of the aforementioned storage media and/or other types of persistent storage. In one or more embodiments of the invention, the persistent storage (150) may be a cloud storage service. A cloud storage service may be a logical storage that includes any number of physical storage devices operable connected and/or unified to form a logical storage. The logical storage may include redundancy or other data integrity features that ensure that failure of any of the physical storage elements does not cause data stored in the logical storage to be lost.

The persistent storage (150) may include an object storage (160) for storing data from the clients (100), a slice recipe storage (170) for storing slice recipes generated by backend micro-services (140) of the deduplication service (110), and an object recipe storage (180) for storing object recipes generated by the frontend micro-services (120). The slice recipe storage (170) and/or the object recipe storage (180) may be stored on different devices and/or different persistent storage without departing from the invention. The slice recipe storage (170) and/or the object recipe storage (180) may be a portion of the object storage (160) without departing from the invention. For additional details regarding the object storage (160), the slice recipe storage (170), and the object recipe storage (180), see FIGS. 2A-2D.

As used herein, an object storage is a data storage architecture that manages data as objects. Each object of the object storage may include data, meta-data, and/or a globally unique identifier of the object. The object may include a number of bytes for storing data in the object. Each object of the object storage may include a different number of bytes without departing from the invention. In one or more embodiments of the invention, the object storage does not include a file system. For additional details regarding the object storage (160), see FIGS. 2A-2B.

The persistent storage (150) may be operably connected to the deduplication service (110). While not shown in FIG. 1, the persistent storage may be operably connected to the clients (100) or other elements of the system without departing from the invention.

The deduplication service (110) may receive data from the clients (100) and deduplicates the data before storing the data in the object storage (160). The deduplication service (110) may also provide data, stored in the object storage, in response to requests for the data from the clients (100). In one or more embodiments of the invention, the deduplication service (110) may be a service executing on a cloud platform, e.g., a platform as a service.

As used herein, a cloud platform is a logical computing resource that abstracts the aggregated physical resources of one or more physical computing systems. One or more of the physical computing systems may be a physical device that includes non-transitory storage, memory (e.g. Random Access Memory), and one or more processors. The non-transitory storage may include instructions which, when executed by the one or more processors, enable the cloud platform to perform the functions described in this application and shown in FIGS. 4-8.

In one or more embodiments of the invention, the deduplication service (110) may be a physical device that includes non-transitory storage, memory (e.g. Random Access Memory), and one or more processors. The physical device may be, for example, a server. The physical device may be other types of computing devices without departing from the invention. The non-transitory storage may include instructions which, when executed by the one or more processors, enable the physical device to perform the functions described in this application and shown in FIGS. 4-8.

The deduplication service (110) may include frontend micro-services (120) that receive data from clients (100), backend micro-services (140) that deduplicate slices of the received data, and a message manager (130) that manages requests and notifications between the frontend micro-services (120) and the backend micro-services (140). The frontend micro-services (120) and backend micro-services (140) may also facilitate providing data stored in the object storage to the clients (100). Each component of the deduplication service (110) is described below.

The frontend micro-services (120) may be one or more services that receive data sent by the clients (100) for storage in the object storage (160), prepare the data for deduplication, and forward the prepared data to the backend micro-services (140). In response to obtaining a data storage request sent by a client, a frontend micro-service may perform the method shown in FIG. 4. In one or more embodiments of the invention, the frontend micro-services (120) may be micro-services executing on a cloud platform.

The frontend micro-services (120) may also obtain requests for data stored in the object storage (160). The frontend micro-services may perform the method shown in FIG. 6 to facilitate reading of data from the object storage (160).

While not illustrated in FIG. 1, the system may include a load balancer that receives all, or a portion thereof, of the requests to store and/or obtain data in the object storage from the clients (100) and distributes the requests to the frontend micro-services. The number of frontend micro-services may be dynamically adjusted, i.e., additional instances of the frontend micro-services may be instantiated or existing instances of the frontend micro-service may be terminated, to match the request processing capacity of the frontend micro-services (120) to the rate of requests for storing data in the object storage from the clients (100). For example, the number of requests from the clients may exceed the request processing capacity of the frontend micro-services (120) resulting in the development of a backlog of requests and thereby result in a mismatch. In another example, the number of requests from the clients may be exceeded by the request processing capacity of the frontend micro-services (120) resulting in idle frontend micro-services and thereby result in a mismatch.

The message manager (130) may facilitate transmission of requests and notifications between the frontend micro-services (120) and the backend micro-services (140). In one or more embodiments of the invention, the message manager (130) may be a service executing on a cloud platform. The message manager (130) may include request queues (131) and notification queues (132). Each of the queues is discussed below.

The request queues (131) may be one or more queues that queue slices of data generated by the frontend micro-services (120) for processing by the backend micro-services (140) or queue slices of data reconstructed by the backend micro-services (140) for use by the frontend micro-services (120) when reconstructing stored data. Each of the queues of the request queues (131) may be first in first out queues. The queues of the request queues (131) may be other types of queues without departing from the invention. For example, the queues may be configured to prioritize certain slices for processing by the backend micro-services (140) over other slices, e.g., certain slices may be moved to the front of the queue based on a type, quality, or meta-data associated with the slices.

In one or more embodiments of the invention, a first queue may be assigned to facilitate storing of data in the object storage and a second queue may be assigned to facilitate reading of data from the object storage. For example, the first queue may send slices of data to the backend micro-services for processing when data is being stored in the object storage and the second queue may send reconstructed slices of data to the frontend micro-services for processing when data is being read from the object storage. The second queue may be a notification queue that enables a backend micro-service to send a reconstructed slice to a specified frontend micro-service. The first queue may be a request queue that enables a frontend micro-service to send a slice request or a slice to any backend micro-service without specifying the specific micro-service that will receive the request. In other words, the first queue may send requests to any backend micro-service while the notification queue may send reconstructed slices to frontend micro-services that requested the slice that has been reconstructed.

The notification queues (132) may be messaging queues that enable the backend micro-services (140) and the frontend micro-services (120) to exchange confirmation of completion messages of any portion of the methods shown in FIGS. 4-7.

The backend micro-services (140) may be one or more micro-services that receive slices of data from the message manager (130) for storage in the object storage (160), deduplicate the slice, and store the deduplicated slice in a compression region in the object storage (160). The backend micro-services may perform the method shown in FIG. 5 to facilitate storing of deduplicated data in the object storage. In one or more embodiments of the invention, the backend micro-services (140) may be micro-services executing on a cloud platform.

The backend micro-services (140) may also obtain requests for slices of data stored in the object storage (160). The backend micro-services may perform the method shown in FIG. 7 to facilitate providing slices of data stored in the object storage.

In one or more embodiments of the invention, the number of backend micro-services may be dynamically adjusted, i.e., additional instances of the backend micro-services may be instantiated or existing instances of the backend micro-service may be terminated, to match the slice processing capacity of the backend micro-services (140) to the rate of requests for storing slices of data and/or retrieving slices of data in the object storage from the clients (100). The number of backend micro-services may be dynamically adjusted by performing the method shown in FIG. 8.

In one or more embodiments of the invention, the backend micro-services and/or frontend micro-services may be adjusted based on a processing load and/or memory usage load of the hardware on which the deduplication service is executing.

The frontend micro-services and backend micro-services may utilize a number of storages to provide the functionality described herein. FIGS. 2A-2D shows diagrams of structures of the storages (160, 170, 180, FIG. 1) that may be utilized by the frontend and/or backend micro-services.

Additionally, while the frontend micro-services and backend micro-services have been described as separate services, embodiments of the invention are not limited to separate services performing the functionality of the frontend and backend micro-services respectively. The functionality of the frontend and backend micro-services may be performed by a single service without departing from the invention. For example, a single service may perform the functionality, described herein, of both the frontend and backend micro-services.

FIG. 2A shows an object storage (160) in accordance with one or more embodiments of the invention. The object storage (160) may store compression regions (161A, 161P) and similarity group meta-data (162A-162Q).

Each of the compression regions (161A, 161P) may store one or more segments of one or more slices of data. As used herein, a compression region is one or more pieces of data that are aggregated and/or compressed.

Each of the similarity group meta-data (162A-162Q) may specify meta-data associated with a similarity group. The meta-data of each similarity group may specify a sketch and a number of fingerprints. The sketch may be used to match slices of data to the similarity group. The fingerprints may be used to determine whether a segment of a slice of data that mapped to a similarity group is already present in the object storage.

FIG. 2B shows an example of a similarity group A meta-data (162A). The similarity group A meta-data (162A) includes a sketch (163) and a number of fingerprints (164A-164R).

In one or more embodiments of the invention, a sketch may be a bit sequence that does not uniquely identify a slice. Rather, the sketch may be a bit sequence that identifies a group of slices that each include similar but unique data or include identical data. In other words, the sketch of multiple, different slices may be the same bit sequence if each slice includes similar but unique data or includes identical data.

In one or more embodiments of the invention, a fingerprint may be a bit sequence that virtually uniquely identifies a segment of a slice. As used herein, virtually uniquely means that the probability of collision between the fingerprints of two segments that specify different data is negligible, compared to the probability of other unavoidable causes of fatal errors. In one or more embodiments of the invention, the probability is 10{circumflex over ( )}-20 or less. In one or more embodiments of the invention, the unavoidable fatal error may be caused by a force of nature such as, for example, a tornado. In other words, the fingerprint of any two segments that specify different data will virtually always be different.

Each fingerprint (164A-164R) of the similarity group A meta-data (162A) may include a compression region identifier (165A-165R). The compression region identifier (165A-165R) may specify a compression region where a segment of a slice having the same fingerprint as specified by the fingerprint (164A-164R) is stored. In other words, each fingerprint (164A-164R) may specify where a segment having the fingerprint specified by the fingerprint (164A-164R) is stored.

While the similarity group meta-data has been illustrated as only including a sketch (163) and fingerprints (164A-164R), the similarity group meta-data may include other data without departing from the invention. For example, the similarity group may specify a length of a compression region, an offset from the start of a compression region, a bit sequence, a name, or other types of data without departing from the invention.

Returning to FIG. 2A, while the object storage (160) has been illustrated as only including compression regions and similarity group meta-data, other types of data may also be stored in the object storage (160) without departing from the invention. For example, the object storage (160) may include the slice recipe storage (170), the object recipe storage (180), meta-data associated with each compression region, or other types of data. The meta-data associated with each compression region may specify segment boundaries between multiple segments that are stored in each compression region, keys to portions of the compression region, or other data that may identify or specify portions of a compression region.

FIG. 2D shows a slice recipe storage (170) in accordance with one or more embodiments of the invention. The slice recipe storage (170) may include a number of slice recipes (171A-171U) that include data that may be used to reconstruct a slice of data. Each slice recipe (171A-171U) may include a similarity group identifier (172A-172U) and fingerprint identifiers (173A-173U).

The slice recipe may have a name, i.e., slice recipe name A, that uniquely identifies the slice of data. Slice recipes may be generated as part of the method of storing the data in the object storage shown in FIG. 5.

The similarity group identifier (172A-172U) may specify a similarity group and the corresponding meta-data (162A-162Q, FIG. 2A) of the similarity group. The similarity group identifier (172A-172U) may be, for example, a key, bit sequence, or other data that enables the corresponding meta-data to be obtained.

The fingerprint identifiers (173A-173U) may specify one or more fingerprints (164A-164R) of the similarity group meta-data that corresponds to the similarity group specified by the similarity group identifier (172A-172U). The fingerprint identifiers (173A-173U) may be, for example, one or more keys, bit sequences, or other data that enables the one or more fingerprints (164A-164R, FIG. 2B) of the similarity group meta-data to be obtained. Obtaining the one or more fingerprints (164A-164R, FIG. 2B) may enable segments of the slice stored in compression regions specified by the compression region identifier (165A-165R, FIG. 2B) to be obtained.

While the slice recipes (171A-171U) have been illustrated as only including a similarity group identifier (172A-172U) and fingerprint identifiers (173A-173U), the slice recipes (171A-171U) may include other data without departing from the invention.

FIG. 2C shows an object recipe storage (180) in accordance with one or more embodiments of the invention. The object recipe storage (180) may include a number of object recipes (181A-181T) that include data that may be used to reconstruct a data object. The data object may be, for example, a data file or a portion of a data file that has been previously stored in the object storage (160, FIG. 2A). Each object recipe (181A-181T) may include a name of an object (182A-182T) and slice identifiers (183A-183T). Object recipes may be generated as part of the method of storing the data in the object storage shown in FIG. 4.

The name of an object (182A-182T) of each object recipe (181A-181T) may be a name of a data object stored in the object storage. The name of the object (182A-182T) may be used to identify the object recipe (181A-181T) when an entity requests to read a data stored in the object storage by the method of reading data shown in FIG. 6.

The slice identifiers (183A-183T) may specify one or more slice recipes (171A-171U) stored in the slice recipe storage (170). The slice recipe identifiers (183A-183T) may be passed to the backend micro-services by the method of reading data shown in FIG. 6 to obtain each of the slices of data corresponding to each of the slice recipes. The slice identifiers (183A-183U) may be, for example, one or more keys, bit sequences, or other data that enables the backend micro-services to obtain a corresponding slice recipe.

While the object recipes (181A-181U) have been illustrated as only including a name of an object (182A-182U) and slice identifiers (183A-183U), the object recipes (181A-181U) may include other data without departing from the invention.

FIGS. 3A-3B show diagrams that illustrate relationships between data (300), slices (310, 320, 330) of the data (300), and segments (311, 312, 313) of a slice (310) of the data (300), respectively. The data (300) may divided into segments and slices as part of the method shown in FIG. 4.

FIG. 3A shows a diagram of data (300) in accordance with one or more embodiments of the invention. The data may be any type of data in any format and of any length.

FIG. 3B shows a diagram of slices (310, 320, 330) of the data (300). Each slice of the data may include separate, distinct portions of the data (300). Each of the slices may be of different, but similar lengths. For example, each slice may include approximately 8 megabytes of data, e.g., a first slice may include 8.05 megabytes of data, the second slice may include 7.93 megabytes of data, etc. In one or more embodiments of the invention, the average amount of data of each slice is between 4 and 16 megabytes.

FIG. 3C shows a diagram of segments (311, 312, 313) of a slice (310) of the data (300). Each segment may include separate, distinct portions of the slice (310). Each of the segments may be of different, but similar lengths. For example, each segments may include approximately 8 kilobytes of data, e.g., a first segment may include 8.03 kilobytes of data, the second segment may include 7.96 kilobytes of data, etc. In one or more embodiments of the invention, the average amount of data of each segment is between 4 and 16 kilobytes.

FIGS. 4-8 show flowcharts in accordance with one or more embodiments of the technology. The methods shown in FIGS. 4 and 5 may facilitate storing data in the object storage while the methods shown in FIGS. 6-7 may facilitate reading stored data from the object storage. The method shown in FIG. 8 may facilitate both reading and storing of data in the object storage by adjusting computing resources dedicated to the reading and/or writing of data to the object storage. While the various steps in these flowcharts are presented and described sequentially, one of ordinary skill in the art will appreciate that some or all of the steps may be executed in different orders, may be combined or omitted, and some or all of the steps may be executed in parallel. Furthermore, the steps may be performed actively or passively.

FIG. 4 shows a flowchart according to one or more embodiments of the invention. The method depicted in FIG. 4 may be used to store data in an object storage in accordance with one or more embodiments of the invention. The method shown in FIG. 4 may be performed by, for example, a frontend micro-service of the deduplication service (110, FIG. 1).

In Step 400, a deduplication service may obtain a data file, or a portion thereof, for storage in an object storage. The data file may be provided to the deduplication service by a client operably connected to the deduplication service.

In one or more embodiments of the invention, the data file may be received by a load balancer of the deduplication service. The load balancer may added the data file to a queue of data files to be stored in the object storage. The load balancer may provide the data file to a frontend micro-service of the deduplication service once the data file reaches the head of the queue.

In one or more embodiments of the invention, the data file may be streamed to the deduplication service. In other words, portions of the data file or a data stream may be sent to the deduplication service overtime. The load balancer may assign a frontend micro-service of the deduplication service to receive the streamed data file or data stream and perform one or more of the other steps of the process shown in FIG. 4 on portions of the streamed data as it is streamed.

In Step 410, the frontend micro-service may divide the data file into multiple segments. The segments may be non-overlapping portions of the data file each having approximately the same length as described with respect to FIG. 3C.

In one or more embodiments of the invention, the data file may be divided into segments by generating a rolling hash of the data file. A rolling hash may be successive hashes of a window as the window moves through the data file. For example, a first hash of the rolling has may be a hash of 64 bytes of data starting at the first byte of the data file, a second hash of the rolling has may be a hash of 64 bytes of data starting at the second byte of the data file, a third hash of the rolling has may be a hash of a 64 bytes of data starting at the third byte of the data file, etc.

A number of segment breakpoints may then be selected by comparing each hash, or a portion thereof, of the rolling hash to a predetermined bit sequence. The starting byte of each hash that matches the predetermined bit pattern may be selected as a segment breakpoint. The location of each selected starting byte may be used as the segment breakpoints. In one or more embodiments of the invention, the predetermined bit sequence may consist of 13 bits.

The segments of the data file may be generated by dividing the data file into portions based on the locations specified by the segment breakpoints. For example, a first segment may begin at the start of the data file and end at the location specified by the first segment break point, a second segment may begin at the location specified by the first segment break point and end at the location specified by the second segment breakpoint, a third segment may begin at the location specified by the second segment break point and end at the location specified by the third segment breakpoint, etc.

In Step 420, the frontend micro-service may group multiple segments into multiple slices. The slices may be non-overlapping portions of the data file each having approximately the same length as described with respect to FIG. 3B.

In one or more embodiments of the invention, the segments may be grouped into slices using the rolling hash of Step 410.

A number of slice breakpoints may be selected by comparing each hash, or a portion thereof, of the rolling hash to a second predetermined bit sequence. The starting byte of each hash that matches the second predetermined bit pattern may be selected as a slice breakpoint. The location of each selected starting byte may be used as the slice breakpoints. In one or more embodiments of the invention, the second predetermined bit sequence may consist of 23 bits.

The slices of the data file may be generated by aggregating the segments based on the locations specified by the slice breakpoints. For example, a first slice may be generated by aggregating the segments that have locations between the start of the data file and the location specified by the first slice break point, a second slice may be generated by aggregating the segments between the location of the first slice break point and the location of the second slice break point, a third slice may be generated by aggregating all of the segments between the location of the second slice break point and the location of the third slice breakpoint, etc.

In Step 430, the frontend micro-service may send each of the slices to one or more backend micro-services.

In one or more embodiments of the invention, the frontend micro-service may send each of the slices by loading the slices into a request queue that queues each of the slices and/or load balances each of the slices across the one or more backend micro-services.

For example, each of the slices may be sequentially loaded into a request queue. The request queue may then provide a slice to a backend micro-service when the micro-service indicates that it is available for processing. The request queue may then provide a second slice to a second backend micro-service when the second micro-service indicates that it is available for processing. The process may be repeated until all of the slices have been provided to backend micro-services. The request queue may send multiple of the slices to the same backend micro-service without departing from the invention.

In one or more embodiments of the invention, the frontend micro-service may wait to receive a notification from one or more backend micro-services that indicates that each slice has been successfully stored before performing steps 440 and/or 450.

In Step 440, the frontend micro-service may generate an object recipe. The object recipe may specify the data shown in, for example, the object recipe A (181A) of FIG. 2C. The recipe may include the name of the object, i.e., the name of the data file that a client will use to reference the data file, and one or more slice identifiers, i.e., data used to identify the slices that make up the data file.

The name of the object may be generated based on a name of the data file provided by the client. In one or more embodiments of the invention, the clients and the frontend micro-services may implement a predetermined naming convention for data files that are stored in the object storage

Each of the slice identifiers may be generated based on a predetermined naming convention between the frontend micro-services and the backend micro-services. For example, the first slice of a data file may be the name of the data file with a slice number, reflecting the relative location of the slice within the data file, appended to the name of the data file.

The slice identifiers may be generated using other methods without departing from the invention. For example, a slice may be given an arbitrary identifier by a frontend micro-service and the slice identifier may be sent to the backend micro-services along with the slice to ensure consistent naming and identification of slices between the frontend micro-services and the backend micro-services.

In Step 450, the frontend micro-service stores the object recipe in an object storage. The object storage may be a storage as illustrated in FIG. 2C. The object storage may be a persistent storage.

FIG. 5 shows a flowchart according to one or more embodiments of the invention. The method depicted in FIG. 5 may be used to store data in an object storage in accordance with one or more embodiments of the invention. The method shown in FIG. 5 may be performed by, for example, a backend micro-service of the deduplication service (110, FIG. 1).

In Step 500, a backend micro-service may obtain a slice of a data file. The slice may be obtained from a message queue that load-balances requests to store slices in the object storage and requests to read slices from the object storage across multiple backend micro-services of the deduplication service. In one or more embodiments of the invention, the slice may be obtained by notifying the messaging queue that the backend micro-service is available to process additional requests. In one or more embodiments of the invention, the backend micro-service may read the slice from the message queue and notify the message manager and/or the message queue that the slice has been read and/or is being processed by the backend micro-service.

In Step 510, the backend micro-service may generate a sketch of the slice. The sketch of the slice may be a sketch as describe with respect to FIG. 2B. The sketch may be obtained by generating a weak hash of the slice.

In one or more embodiments of the invention, the weak hash may include multiple maximal and/or minimal hash values obtained by performing rolling hash functions over the data of the slice. For example, four rolling hash functions may be performed over the data of the slice and maximal hash value seen for each of the four rolling hash functions may be aggregated. A hash of the aggregated hash values may be generated to obtain a single hash. The single hash may be used as the sketch of the slice.

In one or more embodiments of the invention, the sketch is a bit sequence that matches sketches of other slices, already stored in the object storage, that have similar or identical data.

In Step 520, the backend micro-service may match the sketch obtained in step 510 to a sketch specified by one of the similarity group meta-data stored in the object storage (160, FIG. 2A). The sketch may be matched by sequentially comparing the bit sequence of the sketch to the bit sequence of each sketch of each similarity group meta-data stored in the object storage until an identical bit sequence is found.

In one or more embodiments of the invention, a new similarity group meta-data entry is generated and stored in the object storage if the sketch of the slice does not match a sketch of any similarity group meta-data that is already stored in object storage.

In Step 530, the backend micro-service obtains the fingerprints of the similarity group meta-data to which the sketch was matched in step 520. The backend micro-service may obtain the fingerprints by extracting the fingerprints from the object storage and loading them into a memory of a cloud platform, or other computing device, on which the backend micro-service is executing.

In Step 540, the backend micro-service compares each fingerprint of each segment of the slice to each fingerprint obtained in Step 530. For each fingerprint that does not match any of the obtained fingerprints, the segment that corresponds to the fingerprint may be added to one or more compression regions and the matched similarity group meta-data of step 520 may be updated. Updating the similarity group meta-data may include adding a new fingerprint that specifies the compression region identifier of the compression region in which the corresponding segment is stored. The similarity group meta-data may be updated by adding, modifying, or deleting other data from the similarity group meta-data without departing from the invention.

The fingerprint of each segment of the slice may be a bit sequence as described with respect to FIG. 2A. In one or more embodiments of the invention, segment boundaries of each fingerprint of each segment of the slice may be generated using Rabin's fingerprinting algorithm. In one or more embodiments of the invention, each fingerprint of each segment of the slice may be generated using a cryptographic hash function. The cryptographic hash function may be, for example, a message digest (MD) algorithm or a secure hash algorithm (SHA). The message MD algorithm may be MD5. The SHA may be SHA-0, SHA-1, SHA-2, or SHA3. Other fingerprinting algorithms may be used without departing from the invention. In Step 550, the backend micro-service stores the one or more compression regions in the object storage.

In Step 560, the backend micro-service generates a slice recipe that specifies an identifier of the matched similarity group of Step 520 and an identifier of each fingerprint of each segment of the slice of Step 540. The slice recipe may be a slice recipe as illustrated in FIG. 2D.

In Step 570, the backend micro-service may store the slice recipe in a slice recipe storage. The slice recipe storage may be a storage as illustrated in FIG. 2C and may be stored on a persistent storage.

In one or more embodiments of the invention, the backend micro-service may send a storage completion indicator to the frontend micro-service that sent the slice after or concurrently with storing the slice in the object storage.

FIG. 6 shows a flowchart according to one or more embodiments of the invention. The method depicted in FIG. 6 may be used to obtain data from an object storage in accordance with one or more embodiments of the invention. The method shown in FIG. 6 may be performed by, for example, a frontend micro-service of the deduplication service (110, FIG. 1).

In Step 600, a frontend micro-service may obtain a request for a data file stored in an object storage from an entity. The request may be obtained from a load balancer that distributes requests from clients to the frontend micro-service.

In Step 610, the frontend micro-service may obtain an object recipe specified by the request. The object recipe may be an object recipe as illustrated in FIG. 2C.

In Step 620, the frontend micro-service may send requests for each slice specified by the object recipe to one or more backend micro-services. The slice requests may be sent to the backend micro-services via a message manager that queues the requests and, when a backend micro-service is available, distributes the requests to backend micro-services.

In Step 630, the frontend micro-service may receive each slice specified by the object recipe from the one or more backend micro-services. The slices may be received via the message manager via a queue.

In Step 640, the frontend micro-service assembles a data file, or a portion thereof, using the slices obtained in Step 630 and the object recipe obtained in Step 600. The data file may be assembled by appending each of the slices in an order specified by the object recipe.

In Step 650, the frontend micro-service sends the assembled data file to the entity that requested the data file.

FIG. 7 shows a flowchart according to one or more embodiments of the invention. The method depicted in FIG. 7 may be used to obtain data stored in an object storage in accordance with one or more embodiments of the invention. The method shown in FIG. 7 may be performed by, for example, a backend micro-service of the deduplication service (110, FIG. 1).

In Step 700, a backend micro-service may obtain a request for a slice of a data file. The request for the slice may be obtained from a message queue that load-balances requests to store slices in the object storage and requests to read slices from the object storage across multiple backend micro-services of the deduplication service. In one or more embodiments of the invention, the request for the slice may be obtained by notifying the messaging queue that the backend micro-service is available to process additional requests. The message queue may provide the backend micro-service with the request in response to the notification. In one or more embodiments of the invention, the backend micro-service may read the request for the slice from the message queue and notify the message manager and/or the message queue that the request for the slice has been read and/or is being processed by the backend micro-service.

In Step 710, the backend micro-service may obtain a slice recipe associated with the slice from a slice recipe storage. The slice recipe may be a slice recipe as illustrated in FIG. 2D. The slice recipe may be obtained by matching a name of the slice or data included in the request to a slice recipe name of a slice recipe stored in the slice recipe storage. The slice recipe may be obtained by reading the matched slice recipe from the slice recipe storage.

In Step 720, the backend micro-service may obtain similarity group meta-data specified by the slice recipe using a similarity group meta-data identified specified by the slice recipe of step 710.

In Step 730, the backend micro-service may identify compression regions stored in the object storage and specified by the obtained similarity group meta-data.

In Step 740, the backend micro-service may read each of the identified compression regions of step 730 from the object storage to obtain segments of the slice.

In Step 750, the backend micro-service may assemble the slice using the obtained segments of the slice and the slice recipe.

In Step 760, the backend micro-service may send the assembled slice to the frontend micro-service that requested the slice.

FIG. 8 shows a flowchart in accordance with one or more embodiments of the invention. The method depicted in FIG. 8 may be used to modify the quantity of resources dedicated to managing data of an object storage in accordance with one or more embodiments of the invention. The method shown in FIG. 8 may be performed by, for example, a deduplication service (110, FIG. 1).

In Step 800, the number of requests in a request queue between a number of front end micro-services and a number of backend micro-services may be monitored.

In Step 810, the number of frontend micro-services and/or backend micro-services may be adjusted. In other words, new instances of micro-services may be added or executing instances of micro-services may be terminated.

In one or more embodiments of the invention, new instances of backend micro-services may be added if the number of requests in the request queue exceeds a predetermined number. In other words, new instances of backend micro-services may be added if the rate of request processing of the currently executing backend micro-services is less than a rate of requests being added to the queue.

In one or more embodiments of the invention, currently executing instances of backend micro-services may be terminated if the number of requests in the request queue is less than a predetermined number. In other words, currently executing instances of backend micro-services may be terminated if the rate of request processing capacity of the currently executing backend micro-services is greater than a rate of requests being added to the queue.

In one or more embodiments of the invention, new instances of frontend micro-services may be added if the number of requests in the load balancer queue is more than a predetermined number. In other words, new instances of frontend micro-services may be added if the rate of client requests is greater than a rate of requests being handled by front-end micro-services.

In one or more embodiments of the invention, currently executing instances of frontend micro-services may be terminated if the number of requests in the load-balancer queue is less than a predetermined number. In other words, currently executing instances of frontend micro-services may be terminated if the rate of client requests is less than a rate of requests that could be handled by the current set of front-ends micro-services.

In one or more embodiments of the invention, the quantity of frontend and/or backend micro-services may be adjusted based on an aggregate resource utilization of a computing resources such as CPU, memory, and/or network bandwidth. The quantity of frontend and/or backend micro-services may be adjusted when the aggregate usage of a resource exceeds a threshold. The threshold may be, for example, an upper threshold or a lower threshold, respectively, which triggers adding or terminating instances of frontend and/or backend micro-services.

One or more embodiments of the invention may enable one or more of the following: i) separation of preparation of data files for deduplication and deduplication of the prepared files into separate micro-services, ii) dynamic adjustments of computing resources used for preparation of files for deduplication or deduplication of prepared files by adding new instances of micro-services and/or terminating existing instances of micro-services, iii) scaling of deduplication of data across an arbitrary number of computing resources, and iv) minimizing the computing resource cost of deduplicating data for storage in an object storage by only loading a subset of all of the fingerprints of every segment stored in an object storage by performing similarity group mapping.

While the invention has been described above with respect to a limited number of embodiments, those skilled in the art, having the benefit of this disclosure, will appreciate that other embodiments can be devised which do not depart from the scope of the invention as disclosed herein. Accordingly, the scope of the invention should be limited only by the attached claims.

Claims

1. A system for managing an object storage, comprising: a physical device configured to execute first instructions adapted to cause a plurality of frontend micro-services to: obtain a request to store data in an object storage;divide the data into slices; anda second physical device configured to execute second instructions adapted to cause a plurality of backend micro-services to: generate a sketch of each slice;match each slice to a similarity group of a plurality of similarity groups using the respective sketch of each slice;obtain meta-data associated with each matched similarity group;add at least a portion of a slice of the slices to a compression region using the meta-data,wherein the sketch of each slice comprises a bit string that does not uniquely identify the slice.

2. The system of claim 1, further comprising: a request queue that passes the slices from the frontend micro-services to the backend micro-services,wherein a number of available backend micro-services is dynamically adjusted based on a number of pending slices in the request queue.

3. The system of claim 1, further comprising: a load balancer to: obtain requests to store data in the object storage;distribute the requests to the plurality of frontend micro-services,wherein the system dynamically adjusts a number of available frontend micro-services based on a number of pending requests in the load balancer.

4. The system of claim 1, wherein dividing the data into slices comprises: dividing a portion of the data into a plurality of segments;grouping a portion of the segments of the plurality of segments into a slice;generating an object recipe that specifies, in part, the slice; andstoring the object recipe in an object recipe storage.

5. The system of claim 4, wherein dividing the portion of the data into the plurality of segments comprises: generating a rolling hash of the data;obtaining a plurality of segment break points by comparing each hash of the rolling hash to a predetermined bit pattern; anddividing the data based on the plurality of segment break points.

6. The system of claim 5, wherein grouping the portion of the segments of the plurality of segments into the slice comprises: determining a plurality of slice break points by comparing each hash of the rolling hash to a second predetermined bit pattern;identifying the portion of the segments of the plurality of segments based on plurality of slice break points; andgrouping the segments of the identified portion of the segments of the plurality of segments into the slice,wherein a second predetermined bit pattern has a length that is longer than a length of the predetermined bit pattern.

7. The system of claim 1, wherein storing at least the portion of the slice of the slices in the compression region using the meta-data comprises: identifying a portion of the segments of the slice that each have a fingerprint that does not match any fingerprint specified by the meta-data;adding a segment of the identified portion of the segments of the slice to the compression region;updating the meta-data based on the identified portion of the segments;generating a slice recipe that specifies: the matched similarity group, andeach fingerprint of each segment of the slice;storing the slice recipe in a slice recipe storage.

8. The system of claim 7, wherein the fingerprint of each segment comprises a bit string that virtually uniquely identifies the segment.

9. The system of claim 8, wherein the fingerprint of each segment is obtained by performing a secure hash of the segment.

10. The system of claim 1, wherein the plurality of backend services is further to: store the compression region in the object storage,wherein the object storage is a persistent storage.

11. The system of claim 1, wherein the request to store data in the object storage is received from a computing device.

12. The system of claim 11, wherein the plurality of frontend micro-services executes on a cloud platform operably connected to the computing device.

13. The system of claim 12, wherein the plurality of backend micro-services executes on the cloud platform.

14. A method of managing an object storage, comprising: obtaining, by a frontend micro-service of a plurality of frontend micro-services, a request to store data in an object storage;dividing, by the frontend micro-service, the data into slices;generating, by a backend micro-service of a plurality of backend micro-services, a sketch of each slice;matching, by the backend micro-service, each slice to a similarity group of a plurality of similarity groups using the respective sketch of each slice;obtaining, by the backend micro-service, meta-data associated with each matched similarity group; andadding, by the backend micro-service, at least a portion of a slice of the slices to a compression region using the meta-data,wherein the sketch of each slice comprises a bit string that does not uniquely identify the slice.

15. The method of claim 14, further comprising: monitoring a number of pending slices in a request queue that passes the slices from the plurality of frontend micro-services to the plurality of backend micro-services; anddynamically adjusting a number of backend micro-services of the plurality of backend micro-services based on the monitoring the number of pending slices in the request queue.

16. The method of claim 14, further comprising: distributing, by a load balancer, requests to store data in the object storage to the plurality of frontend micro-services;monitoring, by the load balancer, a number of requests of the requests that are queued to be distributed; anddynamically adjusting a number of available frontend micro-services based on the monitoring of the number of requests of the requests that are queued to be distributed.

17. A non-transitory computer readable medium comprising computer readable program code, which when executed by a computer processor enables the computer processor to perform a method for managing an object storage, the method comprising: obtaining, by a frontend micro-service of a plurality of frontend micro-services, a request to store data in an object storage;dividing, by the frontend micro-service, the data into slices;generating, by a backend micro-service of a plurality of backend micro-services, a sketch of each slice;matching, by the backend micro-service, each slice to a similarity group of a plurality of similarity groups using the respective sketch of each slice;obtaining, by the backend micro-service, meta-data associated with each matched similarity group; andadding, by the backend micro-service, at least a portion of a slice of the slices to a compression region using the meta-data,wherein the sketch of each slice comprises a bit string that does not uniquely identify the slice.

18. The non-transitory computer readable medium of claim 17, wherein the method further comprises: monitoring a number of pending slices in a request queue that passes the slices from the plurality of frontend micro-services to the plurality of backend micro-services; anddynamically adjusting a number of backend micro-services of the plurality of backend micro-services based on the monitoring the number of pending slices in the request queue.

19. The non-transitory computer readable medium of claim 17, further comprising: distributing, by a load balancer, requests to store data in the object storage to the plurality of frontend micro-services;monitoring, by the load balancer, a number of requests of the requests that are queued to be distributed; anddynamically adjusting a number of frontend micro-services of the plurality of frontend micro-services based on the monitoring of the number of requests of the requests that are queued to be distributed.