(19)
(11)EP 2 972 744 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
14.08.2019 Bulletin 2019/33

(21)Application number: 14712383.0

(22)Date of filing:  25.02.2014
(51)Int. Cl.: 
G06F 3/06  (2006.01)
G06F 9/46  (2006.01)
(86)International application number:
PCT/US2014/018265
(87)International publication number:
WO 2014/158563 (02.10.2014 Gazette  2014/40)

(54)

OBJECT-BASED TRANSACTIONS IN A STORAGE SYSTEM

OBJEKTBASIERTE TRANSAKTIONEN IN EINEM SPEICHERSYSTEM

TRANSACTIONS BASÉES OBJET DANS UN SYSTÈME DE STOCKAGE


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 14.03.2013 US 201313826792

(43)Date of publication of application:
20.01.2016 Bulletin 2016/03

(73)Proprietor: EMC IP Holding Company LLC
Hopkinton, MA 01748 (US)

(72)Inventor:
  • SHAPIRO, Michael, W.
    Menlo Park, CA 94025 (US)

(74)Representative: Sharman, Thomas Alexander et al
Reddie & Grose LLP The White Chapel Building 10 Whitechapel High Street
London E1 8QS
London E1 8QS (GB)


(56)References cited: : 
US-B1- 7 617 180
US-B1- 8 370 567
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    BACKGROUND



    [0001] Applications typically require various datum to be atomically committed in a transaction. In an object storage system, a transaction is a set of logical storage locations (offsets) corresponding to an object that is to be modified, and new data to be written at each offset, where all of the new data will be visible, or none of it will visible if a failure occurs, but no other intermediate state with only a portion of the modifications is permitted. Traditional mechanisms for implementing transactions typically require the data that is part of the transaction to be written multiple times in order to commit the transaction resulting in a large processing overhead. For example, patent document US 8 370 567 B1 describes a system for writing data to a persistent storage together with related table of content entries.

    SUMMARY



    [0002] In general, in one aspect, the invention relates to a non-transitory computer readable medium comprising instructions, which when executed by a processor, perform a method, the method comprising: creating a transaction object for a transaction, wherein the transaction object is identified by a transaction object ID (TOI) and associated with an object identified by an object ID (OID), storing, in persistent storage, a table of contents entry (TE) and a metadata (MD) frag for the transaction object, wherein TE references the MD frag and the MD frag identifies the object, receiving a write request to write data to the transaction object as part of the transaction, wherein the write request specifies the TOI and an offset, in response to the write request: storing, in the persistent storage, a second TE and a data frag, wherein the second TE specifies the TOI and offset, and wherein the second TE references the data frag and the data frag comprises the data, storing, in an in-memory data structure, an entry comprising a hash value and a physical address of the data frag, wherein the hash value is derived from the TOI and the offset, receiving a commit request to commit the transaction, in response to the commit request:

    [0003] storing, in persistent storage, a third TE and a second metadata (MD) frag for the transaction object, wherein third TE references the second MD frag, wherein the second MD frag identifies the object and specifies that the transaction is committed, updating, in the in-memory data structure, a second entry comprising a second hash value and a second physical address for a second data frag to replace the second physical address with the physical address, wherein the second hash value is derived from the OID and the offset.

    [0004] The invention is defined in its various aspects in the independent and dependent claims.

    [0005] Embodiments of the invention are described in detail by way of example with reference to the drawings,

    BRIEF DESCRIPTION OF DRAWINGS



    [0006] 

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

    FIG. 2A shows a solid-state memory module in accordance with one or more embodiments of the invention.

    FIG. 2B shows a block in accordance with one or more embodiments of the invention.

    FIG. 2C shows a frag page in accordance with one or more embodiments of the invention.

    FIG. 2D shows a TOC page in accordance with one or more embodiments of the invention.

    FIG. 2E shows a block in accordance with one or more embodiments of the invention.

    FIG. 2F shows a table of contents (TOC) entry in accordance with one or more embodiments of the invention.

    FIG. 2G shows a data frag in accordance with one or more embodiments of the invention.

    FIG. 2H shows a metadata frag in accordance with one or more embodiments of the invention.

    FIG. 2I shows a compound frag in accordance with one or more embodiments of the invention.

    FIG. 3 shows an object and corresponding TOC entries in accordance with one or more embodiments of the invention.

    FIG. 4 shows a flowchart for performing a single object transaction in accordance with one or more embodiments of the invention.

    FIGS. 5A-5C show a flowchart for performing a multi-object transaction in accordance with one or more embodiments of the invention.

    FIGS. 6A-6D show an example in accordance with one or more embodiments of the invention.

    FIGS. 7A-7D show an example in accordance with one or more embodiments of the invention.


    DETAILED DESCRIPTION



    [0007] Specific embodiments of the invention will now be described in detail with reference to the accompanying figures. In the following detailed description of embodiments of the invention, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to one of ordinary skill in the art that the invention may be practiced without these specific details. In other instances, well-known features have not been described in detail to avoid unnecessarily complicating the description.

    [0008] In general, aspects of the invention relate to a method and system for implementing transactions in a storage system. More specifically, embodiments of the invention use transaction objects to implement the transaction, where each transaction only requires updated datum/data to be written once to the persistent storage. In one embodiment of the invention, the transactions may be associated with a single object (see e.g., FIG. 4 and 6A-6D) or multiple objects (see e.g., FIGS. 5A-5C and 7A-7D). Further, embodiments of the invention allow a single copy of the data to represent the transaction during all intermediate states and following a transaction commit.

    [0009] FIG. 1 shows a system in accordance with one embodiment of the invention. As shown in FIG. 1, the system includes one or more clients (100), a storage controller (102), memory (104), and a storage array (106).

    [0010] In one embodiment of the invention, a client (100) is any system or process executing on a system that includes functionality to issue a read request, a write request, or a commit request (with one or more transaction objects) to the storage controller (102). In one embodiment of the invention, the clients (100) may each include a processor (not shown), memory (not shown), and persistent storage (not shown). In one embodiment of the invention, the storage controller (102) is operatively connected to one or more clients (100) and is configured to implement the transaction object scheme, which includes writing table of contents entries and frags to the storage array in a manner consistent with the transaction object scheme described below (see e.g., FIG. 4-5C). In one embodiment of the invention, the storage controller (102) includes a processor (not shown) configured to execute instructions to implement one or more embodiments of the invention, where the instructions are stored on a non-transitory computer readable medium (not shown) that is located within or that is operatively connected to the storage controller (102). Alternatively, the storage controller (102) may be implemented using hardware.

    [0011] In one embodiment of the invention, the storage controller (102) is operatively connected to memory (104). The memory (104) may be any volatile memory including, but not limited to, Dynamic Random-Access Memory (DRAM), Synchronous DRAM, SDR SDRAM, and DDR SDRAM. In one embodiment of the invention, the memory (104) is configured to temporarily store various data (including data for table of contents entries and frags) prior to such data being stored in the storage array.

    [0012] In one embodiment of the invention, the storage array (106) includes a number of individual persistent storage devices including, but not limited to, magnetic memory devices, optical memory devices, solid-state memory devices, phase change memory devices, any other suitable type of persistent memory device, or any combination thereof.

    [0013] System configurations other than the one shown in FIG. 1 may be used.

    [0014] As described above, the storage array may include solid-state memory devices. The following discussion describes embodiments of the invention implemented using solid-state memory devices. Turning to FIG. 2A, FIG. 2A shows a solid-state memory module of a solid-state memory device (not shown) in accordance with one or more embodiments of the invention. The solid-state memory device (not shown) may include multiple solid-state memory modules. The solid-state memory module (200) includes one or more blocks. In one embodiment of the invention, a block is the smallest erasable unit of storage within the solid-state memory module (200).

    [0015] FIG. 2B shows a block in accordance with one or more embodiments of the invention. More specifically, each block (202) includes one or more pages. In one embodiment of the invention, a page is the smallest addressable unit for read and program operations (including the initial writing to a page) in the solid-state memory module. In one embodiment of the invention, rewriting a page within a block requires the entire block to be rewritten. In one embodiment of the invention, each page within a block is either a Frag Page (see FIG. 2C) or a TOC Page (see FIG. 2D).

    [0016] FIG. 2C shows a frag page in accordance with one or more embodiments of the invention. In one embodiment of the invention, the frag page (204) includes one or more frags. In one embodiment of the invention, a frag corresponds to a finite amount of user data (or metadata if the frag page is a metadata frag (see FIG. 2H) or TOC entries if the frag is a compound frag (see FIG. 2I)). Further, the frags within a given page may be of a uniform size or of a non-uniform size. Further, frags within a given block may be of a uniform size or of a non-uniform size. In one embodiment of the invention, a given frag may be less than the size of a page, may be exactly the size of a page, or may extend over one or more pages. In one embodiment of the invention, a frag page only includes frags. In one embodiment of the invention, each frag includes user data (i.e., data provided by the client for storage in the storage appliance) (see FIG. 2G), metadata for an object (see FIG. 2H) or TOC entries, if the frag is a compound frag (see FIG. 2I). A given frag page may include a frags of a single type or frags of two or more types.

    [0017] FIG. 2D shows a TOC page in accordance with one or more embodiments of the invention. In one embodiment of the invention, the TOC page (206) includes one or more TOC entries, where each of the TOC entries includes metadata for a given frag. In addition, the TOC page (206) may include a reference to another TOC page in the block (202). In one embodiment of the invention, a TOC page only includes TOC entries (and, optionally, a reference to another TOC page in the block), but does not include any frags. In one embodiment of the invention, each TOC entry corresponds to a frag (see FIG. 2C) in the block (202). The TOC entries only correspond to frags within the block. Said another way, the TOC page is associated with a block and only includes TOC entries for frags in that block. In one embodiment of the invention, the last page that is not defective in each block within each of the solid-state memory modules is a TOC page.

    [0018] FIG. 2E shows a block in accordance with one or more embodiments of the invention. More specifically, FIG. 2E shows a block (208) that includes TOC pages (210, 212, 214) and frag pages (216, 218, 220, 222, 224, 226). In one embodiment of the invention, the block (208) is conceptually filled from "top" to "bottom." Further, TOC pages are generated and stored once the accumulated size of the TOC entries for the frags in the frag pages equal the size of a page. Turning to FIG. 2E, for example, frag page 0 (216) and frag page 1 (218) are stored in the block (208). The corresponding TOC entries (not shown) for the frags (not shown) in frag page 0 (216) and frag page 1 (218) have a total cumulative size equal to the size of a page in the block. Accordingly, a TOC page (212) is generated (using the TOC entries corresponding to frags in the block) and stored in the block (208). Frag page 2 (220) is subsequently written to the block (208). Because the TOC entries corresponding to the frags (not shown) in frag page 2 (220) have a total cumulative size equal to the size of a page in the block, TOC page (212) is created and stored in the block (208). Further, because there is already a TOC page in the block (208), TOC page (212) also includes a reference to TOC page (214).

    [0019] This process is repeated until there is only one page remaining in the block (208) to fill. At this point, a TOC page (210) is created and stored in the last page of the block (208). Those skilled in the art will appreciate that the total cumulative size of the TOC entries in the TOC page (210) may be less than the size of the page. In such cases, the TOC page may include padding to address the difference between the cumulative size of the TOC entries and the page size. Finally, because there are other TOC pages in the block (208), TOC page (210) includes a reference to one other TOC page (212).

    [0020] As shown in FIG. 2E, the TOC pages are linked from the "bottom" of the block to "top" of the page, such that the TOC page may be obtained by following a reference from a TOC page that is "below" the TOC page. For example, TOC page (212) may be accessed using the reference in TOC page (210).

    [0021] Those skilled in the art will appreciate that while block (208) only includes frag pages and TOC pages, block (208) may include pages (e.g., a page that includes parity data) other than frag pages and TOC pages without departing from the invention. Such other pages may be located within the block and, depending on the implementation, interleaved between the TOC pages and the frag pages.

    [0022] FIG. 2F shows a TOC entry in accordance with one or more embodiments of the invention. In one embodiment of the invention, each TOC entry (230) includes metadata for a frag and may include one or more of the following fields: (i) object ID (232), which identifies the object being stored; (ii) the birth time (234), which specifies the time (e.g., the processor clock value of the processor in the control module) at which the frag corresponding to the TOC entry was written to the storage array; (iii) offset ID (236), which identifies a point in the object relative to the beginning of the object (identified by the object ID); (iv) fragment size (238), which specifies the size of the frag; (v) page ID (240), which identifies the page in the block in which the frag is stored; (vi) byte (242), which identifies the starting location of the frag in the page (identified by the page ID); (vii) logical length (244), which specifies the non-compressed length of the user data in the frag; and (viii) type (246), which specifies the type of data in the frag (e.g., object, compound frag, etc.). The TOC entry (230) may include other fields (denoted by other (248)) without departing from the invention.

    [0023] In one embodiment of the invention, the <object ID, offset ID> or <object ID, offset ID, birth time> identify data, for example, that is provided by the client. Further, the <object ID, offset ID> or <object ID, offset ID, birth time> may be used by the client to identify particular data, while the storage appliance uses a physical address(es) to identify data within the storage appliance. Those skilled in the art will appreciate that the client may provide a logical address instead of the object ID and offset ID.

    [0024] Those skilled in the art will appreciate that the TOC entry may include additional or fewer fields than shown in FIG. 2F without departing from the invention. Further, the fields in the TOC entry may be arranged in a different order and/or combined without departing from the invention. In addition, while the fields in the TOC entry shown in FIG. 2F appear to all be of the same size, the size of various fields in the TOC entry may be non-uniform, with the size of any given field varying based on the implementation of the TOC entry.

    [0025] FIG. 2G shows a data frag in accordance with one or more embodiments of the invention. The data frag (250) only includes data from the client. For example, the data in the data frag may correspond to data in a file, where the file is identified by the object ID.

    [0026] FIG. 2H shows a metadata frag in accordance with one or more embodiments of the invention. The metadata frag (252) includes object metadata for the object corresponding to the object ID in the TOC entry associated with the metadata frag. Said another way, the metadata for the object is stored in a metadata frag and associated with a TOC entry that includes the object ID for the object and offset = -1. Values other than -1 may be used. Examples of object metadata may include, but are not limited to, the object type, its maximum size, its creation time, its most recent modification time, and other per-object attributes. In one embodiment of the invention, there is one metadata frag for each object in the storage array.

    [0027] FIG. 2I shows a compound frag in accordance with one or more embodiments of the invention. In one embodiment of the invention, the compound frag (254) unlike other frags does not include data from a client or metadata for an object; rather, the compound frag (254) includes two or more TOC entries, where each of the TOC entries is associated with a transaction object and has an offset = -1. In one embodiment of the invention, all transaction objects specified in the TOC entries in the compound frag belong to a single transaction.

    [0028] FIG. 3 shows an object (300) and corresponding TOC entries in accordance with one or more embodiments of the invention. More specifically, FIG. 3 shows the conceptual view of an object and corresponding TOC entries in accordance with one embodiment of the invention. As shown in FIG. 3, the object (300) includes a metadata frag located at offset -1, which includes metadata for the object. Further, the object includes two pieces of data (stored in one or more frag pages, not shown) located at offsets 3 and 6. Each of the aforementioned frags is associated with a TOC entry (stored in one or more TOC pages, not shown).

    [0029] FIGS. 4-5C show flowcharts in accordance with one or more aspects of the invention. While the various steps in the flowchart are presented and described sequentially, one of ordinary skill 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. In one aspect of the invention, the steps shown in FIGS. 4-5C may be performed in parallel with any other steps shown in FIGS. 4-5C.

    [0030] Turning to FIG. 4, FIG. 4 shows a flowchart for performing a single object transaction in accordance with one or more embodiments of the invention. In Step 400, a request for a transaction is received by the storage controller, where the request specifies a parent object. The parent object corresponds to the object on which the transaction is to be performed. (See e.g., FIG. 6A, parent object is object with object ID =A).

    [0031] In Step 402, a transaction object is created for the transaction. Creating the transaction objection may include specifying an object ID for the transaction. In Step 404, a metadata (MD) frag is created for the transaction object. The metadata frag may include the transaction object ID (TOI), the object ID for the parent object, and a field indicating that the transaction is not committed. The metadata frag may also include other metadata for the transaction object.

    [0032] In Step 406, a table of contents (TOC) entry (referred to as MD TE) is created for the metadata frag. The MD TE includes the TOI and an offset = -1. The MD TE may include other information without departing from the invention. In Step 408, the MD frag and MD TE are subsequently stored in the storage array. In one embodiment of the invention, the MD frag is stored in the storage array as part of a frag page and the MD TE is stored in the storage array as part of a TOC page. The timing of the writing aforementioned frags, TEs and pages, may vary based on how embodiments of the invention are implemented.

    [0033] In Step 410, the TOI is provided to the client that initiated the transaction. The TOI may be provided to other clients that are participating in the transaction without departing from the invention. In Step 412, a write request to write data as part of the transaction is received. The write request includes the TOI and an offset.

    [0034] In Step 414, a data frag with the data identified in the write request is stored in the storage array (typically as part of a frag page). In Step 416, a TE for the data frag (stored in Step 414) is created and stored in the storage array (typically as part of a TOC page). The TE for the data frag includes the TOI and the offset. The TE does not include the object ID for the parent object.

    [0035] In Step 418, an in-memory data structure is updated to include an entry that includes (i) a hash value derived from the TOI and offset and (ii) the physical address of the data frag stored in step 414. In one embodiment of the invention, the hash value may be derived by applying a function (e.g., MD5, SHA 1, etc.) to <TOI, offset> (or to a value generated using TOI and offset). Those skilled in the art will appreciate that any injective function may be used without departing from the invention. Returning to FIG. 4, at this stage the storage controller can process a read request that includes the TOI and the offset and a read request that includes the object ID of the parent object and the offset. The results of servicing the aforementioned read requests will typically result in obtaining two different datum - one corresponding to <TOI, Offset> and one corresponding to <Parent Object ID, offset>, where the offsets are the same.

    [0036] In Step 420, a determination is made about whether there are additional write requests for the transaction. If there are additional write requests for the transaction the process proceeds to Step 412; otherwise, the process proceeds to step 422. In Step 422, a commit request specifying the TOI is received from the client. In Step 424, a second MD frag is created for the transaction object that includes the transaction object ID (TOI), the object ID for the parent object, and a field indicating that the transaction is committed.

    [0037] In Step 426, a second MD TE is created where the second MD TE includes the TOI and offset = -1. Steps 424 and 426 are performed when the storage array and/or storage controller implements a copy-on-write scheme. If the storage array and/or storage controller do not implement a copy-on-write scheme, then steps 424 and 426 may not be performed; rather the MD TE and MD frag created in steps 404 and 406 may be updated to include the information to include the same information as the MD TE and MD frag created in Steps 424 and 426, respectively.

    [0038] In Step 428, the second MD frag is stored in the storage array (typically in a frag page) and the second MD TE is stored in the storage array (typically in a TOC page). In Step 430, the in-memory data structure is updated to reflect that the transaction is committed. Specifically, in one embodiment of the invention, an entry that includes a hash value derived from the parent object ID and the offset is updated to include the physical address of data, where the physical address was previously present in an another entry (i.e., an entry created in Step 418) that includes the TOI and the offset. If more than one write request was received and processed in the transaction (see Steps, 412-420), then multiple entries in the in-memory data structure are updated. At this stage, the controller can service read requests using <parent object ID, offset> but cannot service any requests specifying the TOI. Further, when a read request including <parent object ID, offset> is received where the offset corresponds to an offset received in a write request in Step 412, the result is the data that was stored in the data frag in step 414.

    [0039] Turning to FIGS. 5A-5C, FIGS. 5A-5C show a flowchart for performing a multi-object transaction in accordance with one or more aspects of the invention. In one aspect of the invention, FIGS. 5A-5C show a method for implementing transactions that are related to multiple objects, where each object is associated with its own transaction object. As described below, each object and corresponding transaction object may operate independently of any other object and corresponding transaction object until the overall transaction (that includes multiple objects) is committed. Accordingly, in the discussion below, a "transaction" may refer to either a transaction that involves a single transaction object or a transaction that involves multiple transaction objects.

    [0040] Turning to FIG. 5A, in Step 500, a request for a transaction is received by the storage controller, where the request specifies a parent object. The parent object corresponds to the object on which the transaction is to be performed. (See e.g., FIG. 7A, parent object is object with object ID =A). In Step 502, a transaction object is created for the transaction. Creating the transaction object may include specifying an object ID for the transaction. In Step 504, a metadata (MD) frag is created for the transaction object. The metadata frag may include the transaction object ID (TOI), the object ID for the parent object, and a field indicating that the transaction is not committed. The metadata frag may also include other metadata for the transaction object.

    [0041] In Step 506, a table of contents (TOC) entry (referred to as MD TE) is created for the metadata frag. The MD TE includes the TOI and an offset = -1. The MD TE may include other information without departing from the invention. In Step 508, the MD frag and MD TE are subsequently stored in the storage array. In one aspect of the invention, the MD frag is stored in the storage array as part of a frag page and the MD TE is stored in the storage array as part of a TOC page. The timing of the writing of the aforementioned frags, TEs and pages, may vary based on how aspects of the invention are implemented.

    [0042] In Step 510, the TOI is provided to the client that initiated the transaction. The TOI may be provided to other clients that are participating in the transaction. In Step 512, a determination is made about whether there are additional transactions to create. If so, the process proceeds to step 502; otherwise the process proceeds to step 514. In Step 514, a write request to write data as part of one of the transactions is received. The write request includes the TOI and an offset. The TOI may correspond to any TOI created during an iteration of step 502. In Step 516, a data frag with the data identified in the write request is stored in a data frag in the storage array (typically as part of a frag page). In Step 518, a TE for the data frag (stored in Step 516) is created and stored in the storage array (typically as part of a TOC page). The TE for the data frag includes the TOI and the offset. The TE does not include the object ID for the parent object.

    [0043] In Step 520, an in-memory data structure is updated to include an entry that includes (i) a hash value derived (as discussed above in FIG. 4) from the TOI and offset and (ii) the physical address of the data frag stored in step 516. At this stage the storage controller can process a read request that includes the TOI and the offset and a read request that includes the object ID of the parent object and the offset. The results of servicing the aforementioned read requests will typically result in obtaining two different datum - one corresponding to <TOI, Offset> and one corresponding to <Parent Object ID, offset>, where the offsets are the same. In Step 522, a determination is made about whether there are additional write requests for any of the transactions (i.e., write requests that specify one of the TOIs created during any iteration of Step 502). If there are additional write requests for any of the transactions the process proceeds to Step 514; otherwise, the process proceeds to step 524.

    [0044] In Step 524, a commit request specifying two or more TOIs is received from the client(s). The TOIs correspond to the transactions created in Step 502. In Step 526, a second MD frag is created for each transaction object (i.e., transaction objects corresponding to transactions created in Step 502) that includes the corresponding transaction object ID (TOI), the corresponding object ID for the parent object, and a field indicating that the transaction is committed. In addition, a second MD TE is created for each of the aforementioned second MD frags, where the second MD TEs include the corresponding TOI and offset = -1. Step 526 is performed when the storage array and/or storage controller implements a copy-on-write scheme. If the storage array and/or storage controller do not implement a copy-on-write scheme, step 526 may not be performed; rather the MD TE and MD frag created in steps 504 and 506 may be updated to include the information to include the same information as the MD TE and MD frag created in Steps 424 and 426, respectively.

    [0045] In Step 528, a compound frag is created where the compound frag is associated with a compound object and includes the MD TEs created in Step 526. In Step 530, a TE for the compound frag is created. The TE for the compound frag includes an object ID for the compound frag and an offset = -1. The TE for the compound frag may also indicate that the frag referenced by the compound TE is a compound frag. In Step 532, the compound frag, the TE for the compound frag, and the MD frags (created in Step 526) are stored in the storage array. The compound frag and MD frags are stored in one or more frag pages and the TE for the compound frag is stored in a TOC page. In one embodiment of the invention, the storing of the compound frag implicitly indicates that all transactions specified in the compound frag (via transaction object IDs) have been committed.

    [0046] In Step 534, the in-memory data structure is updated to reflect that the transactions are committed. Specifically, in one aspect of the invention, multiple entries in the in-memory data structure are updated in the manner discussed above with respect to Step 430 in FIG. 4.

    [0047] FIGS. 6A-6D show an example in accordance with one or more aspect of the invention. The example is not intended to limit the scope of the invention.

    [0048] Referring to FIG. 6A, consider the scenario in which there is an object with object ID = A in the storage array. More specifically, there is a MD TOC entry for the object and a corresponding MD frag for the object, where the MD frag is accessible via hash<A, -1>. Further, there is a data frag (data frag B) that includes data for the object, where the data is associated with <A, 3>. The aforementioned object ID and offset are stored in a corresponding data TE (Data TOC Entry B) in the storage array. At this stage there are four items in the storage array and two entries in the in-memory data structure. Each of the entries in the in-memory data structure points to one of the aforementioned frags.

    [0049] Referring to FIG. 6B, consider a scenario in which a client requests a transaction specifying object ID = B. In this scenario, a MD frag (metadata frag C) is created and stored in the storage array, where the MD frag specifies that the transaction object ID = B, that type of object B is a transaction, that the parent Object ID = A, and that the transaction is not committed. A corresponding MD TE (metadata TOC entry C) is also created and stored in the storage array, where the MD TE specifies object ID = B and offset = -1. A corresponding entry for metadata frag C is also created and stored in the in-memory data structure.

    [0050] Referring to FIG. 6C, consider a scenario in which a write request for the transaction is subsequently received and processed. The result of servicing the write request is a data frag D with data (Data') and a corresponding data TE (data TOC entry D) specifying object ID = B and offset = 3. A corresponding entry for data frag D is also created and stored in the in-memory data structure. At this stage, four items in the storage array may be accessed via the corresponding entries in the in-memory data structure.

    [0051] Referring to FIG. 6D, after the write request is received and processed as described above with respect to FIG. 6C, assume that a commit request including object ID = B is received. At this stage, the transaction is committed. Committing the transaction, as shown in the example, may include (i) updating or creating a new metadata frag C that specifies that the transaction corresponding to object ID = B is committed; (ii) invaliding data frag B and data TOC entry B; and (iii) updating the in-memory data structure to associate hash<A, 3> with the physical address of data frag D.

    [0052] FIGS. 7A-7D show an example in accordance with one or more aspects of the invention. The example is not intended to limit the scope of the invention.

    [0053] Consider the scenario in which there are two transactions one associated with each of the following transaction objects - object ID = B and object ID = D. Further, assume that the parent object ID for transaction associated with object ID = B is object ID = A and that the parent object ID for transaction associated with object ID = D is object ID = C. Finally, assume that there is data (in the form of data frags) associated with each of object ID = A and object ID = C. FIG. 7A shows the content of the storage array based on the above assumptions. Further, FIG. 7B shows the content of the in-memory data structures based on the above assumptions.

    [0054] Assume that a write request is serviced for each of the aforementioned transactions resulting in two additional data TEs (data TOC entry B and data TOC entry D) and two additional data frags (data frag B and data frag D) being stored in the storage array, where data frag B is associated with the same offset as data frag A and data frag D is associated with the same offset as data frag C. In addition, assume that after both the write requests have been serviced, a commit request specifying object ID = B and object ID = D is received. The commit request is subsequently serviced by creating and storing a compound frag and corresponding MD TE (metadata TOC entry E) in the storage array. Finally, data frag A, data TOC entry A, data frag C, and data TOC entry C are invalidated in the storage array; invalidating the corresponding hash entries in the in-memory data structure. FIG. 7C shows the content of the storage array based on the above assumptions. Further, FIG. 7D shows the content of the in-memory data structures based on the above assumptions.

    [0055] In one aspect of the invention, the initial in-memory data structure may be generated by reading in all of the metadata TOC entries and corresponding frags (i.e., metadata frags or compound frags). Once this data is obtained, the storage controller (or a related process) will be able to "replay" the transactions by reviewing the information in the metadata frags and compound frags to re-generate the last state of the in-memory data structure. In particular, if a given MD frag indicates that the object is a transaction (or transaction object) and that the transaction is not committed, then the data frags associated with the transaction (i.e., data frags corresponding with TEs specifying the transaction object ID) are not used to create entries in the in-memory data structure. Further, when a compound frag is identified, the storage controller (or related process) recursively processes the metadata TEs in the compound frag.

    [0056] In one aspect of the invention, the in-memory data structure is generated prior to any operations (e.g., read operation, a write operation, and/or an erase operation) being performed on any datum stored in the solid-state memory modules.

    [0057] Those skilled in the art will appreciate that while the invention has been described with respect to the last page in each block being reserved as a TOC page, aspects of the invention may be implemented by setting another page in the block as a reserved TOC page.

    [0058] While aspects of the invention have been described with respect to an implementation on solid-state memory devices, aspects of the invention may be implemented using another type(s) of storage device(s) without departing from the invention.

    [0059] One or more aspects of the invention may be implemented using instructions executed by one or more processors in the system. Further, such instructions may correspond to computer readable instructions that are stored on one or more non-transitory computer readable mediums.

    [0060] While the invention has been described with respect to a limited number of embodiments, those skilled in the art, having 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 non-transitory computer readable medium comprising instructions, which when executed by a processor, perform a method, the method comprising:

    creating (402) a transaction object for a transaction to be performed for an object, wherein the transaction object is identified by a transaction object ID (TOI) and associated with the object identified by an object ID (OID);

    storing (408), in persistent storage, a first table of contents entry (TE) and a metadata (MD) frag for the transaction object, wherein the first TE references the MD frag, wherein the MD frag corresponds to a finite amount of metadata and identifies the object;

    receiving (412) a write request to write data to the transaction object as part of the transaction, wherein the write request specifies the TOI and an offset;

    in response to the write request:

    storing (414, 416), in the persistent storage, a second TE and a data frag, wherein the second TE specifies the TOI and the offset, and wherein the second TE references the data frag, wherein the data frag corresponds to a finite amount of user data and comprises the data;

    storing (414, 416), in an in-memory data structure located in a memory connected to a storage controller, an entry comprising a hash value and a physical address of the data frag, wherein the hash value is derived from the TOI and the offset;

    receiving (422) a commit request to commit the transaction;

    in response to the commit request:

    storing (428), in persistent storage, a third TE and a second metadata (MD) frag for the transaction object, wherein third TE references the second MD frag, wherein the second MD frag identifies the object and specifies that the transaction is committed;

    updating (430), in the in-memory data structure, a second entry comprising a second hash value and a second physical address for a second data frag to replace the second physical address with the physical address, wherein the second hash value is derived from the OID and the offset;

    wherein prior to the commit request being processed, data in the object is accessible using the OID and the offset and the data in the data frag associated with the transaction, but not committed, is accessible using the TOI and the offset; and

    wherein after the commit request is processed, the data in the data frag is accessible using the OID and the offset and second data in the second data frag is not accessible.


     
    2. The non-transitory computer readable medium of claim 1, wherein the persistent storage is solid-state memory.
     
    3. The non-transitory computer readable medium of claim 1, wherein the commit request specifies the TOI and does not specify the OID.
     
    4. The non-transitory computer readable medium of claim 1, wherein storing the first TE comprises storing the first TE in a table of contents (TOC) page in the persistent storage.
     
    5. The non-transitory computer readable medium of claim 1, wherein the first TE and the MD frag are stored in separate pages in a block in the persistent storage.
     
    6. The non-transitory computer readable medium of claim 1, wherein the MD frag identifies the object using the OID.
     
    7. The non-transitory computer readable medium of claim 1, wherein prior to processing the commit request, the MD frag specifies that the transaction is not committed.
     
    8. The non-transitory computer readable medium of claim 1, wherein the first TE includes a third offset equal to -1.
     


    Ansprüche

    1. Nichtflüchtiges computerlesbares Medium, das Anweisungen aufweist, die bei Ausführung durch einen Prozessor ein Verfahren durchführen, wobei das Verfahren Folgendes aufweist:

    Erstellen (402) eines Transaktionsobjekts für eine für ein Objekt durchzuführende Transaktion, wobei das Transaktionsobjekt durch eine Transaktionsobjektkennung (TOI) gekennzeichnet wird und dem durch eine Objektkennung (OID) gekennzeichneten Objekt zugeordnet wird;

    Speichern (408) eines ersten Inhaltsverzeichnisseintrags (TE) und eines Metadaten- (MD) -fragments für das Transaktionsobjekt in einem persistenten Speicher, wobei der erste TE auf das MD-Fragment verweist, wobei das MD-Fragment einer endlichen Metadatenmenge entspricht und das Objekt kennzeichnet;

    Empfangen (412) einer Schreibanforderung zum Schreiben von Daten in das Transaktionsobjekt als Teil der Transaktion, wobei die Schreibanforderung die TOI und ein Offset angibt;

    als Antwort auf die Schreibanforderung:

    Speichern (414, 416) eines zweiten TE und eines Datenfragments im persistenten Speicher, wobei der zweite TE die TOI und das Offset angibt und wobei der zweite TE auf das Datenfragment verweist, wobei das Datenfragment einer endlichen Nutzerdatenmenge entspricht und die Daten aufweist;

    Speichern (414, 416) eines Eintrags, der einen Hash-Wert und eine physische Adresse des Datenfragments aufweist, wobei der Hash-Wert von der TOI und dem Offset abgeleitet ist, in einer Datenstruktur im Arbeitsspeicher, die sich in einem mit einem Speicher-Controller verbundenen Arbeitsspeicher befindet;

    Empfangen (422) einer Commit-Anforderung zum Abschließen der Transaktion;

    als Antwort auf die Commit-Anforderung:

    Speichern (428) eines dritten TE und eines zweiten Metadaten- (MD) -fragments für das Transaktionsobjekt in persistentem Speicher, wobei der dritte TE auf das zweite MD-Fragment verweist, wobei das zweite MD-Fragment das Objekt kennzeichnet und angibt, dass die Transaktion abgeschlossen wird;

    Aktualisieren (430) eines zweiten Eintrags, der einen zweiten Hash-Wert und eine zweite physische Adresse für ein zweites Datenfragment aufweist, in der Datenstruktur im Arbeitsspeicher, um die zweite physische Adresse durch die physische Adresse zu ersetzen, wobei der zweite Hash-Wert von der OID und dem Offset abgeleitet ist;

    wobei vor der Verarbeitung der Commit-Anforderung unter Verwendung der OID und des Offsets auf Daten im Objekt zugegriffen werden kann und auf die Daten in dem Datenfragment, das der Transaktion zugeordnet, aber nicht abgeschlossen ist, unter Verwendung der TOI und des Offsets zugegriffen werden kann; und

    wobei nach der Verarbeitung der Commit-Anforderung auf die Daten im Datenfragment unter Verwendung der OID und des Offsets zugegriffen werden kann und auf zweite Daten im zweiten Datenfragment nicht zugegriffen werden kann.


     
    2. Nichtflüchtiges computerlesbares Medium nach Anspruch 1, wobei der persistente Speicher ein Halbleiterspeicher ist.
     
    3. Nichtflüchtiges computerlesbares Medium nach Anspruch 1, wobei die Commit-Anforderung die TOI angibt und die OID nicht angibt.
     
    4. Nichtflüchtiges computerlesbares Medium nach Anspruch 1, wobei das Speichern des ersten TE das Speichern des ersten TE in einer Inhaltsverzeichnis- (TOC) -seite im persistenten Speicher aufweist.
     
    5. Nichtflüchtiges computerlesbares Medium nach Anspruch 1, wobei der erste TE und das MD-Fragment in separaten Seiten in einem Block im persistenten Speicher gespeichert werden.
     
    6. Nichtflüchtiges computerlesbares Medium nach Anspruch 1, wobei das MD-Fragment das Objekt unter Verwendung der OID kennzeichnet.
     
    7. Nichtflüchtiges computerlesbares Medium nach Anspruch 1, wobei das MD-Fragment vor der Verarbeitung der Commit-Anforderung angibt, dass die Transaktion nicht abgeschlossen ist.
     
    8. Nichtflüchtiges computerlesbares Medium nach Anspruch 1, wobei der erste TE ein drittes Offset gleich -1 enthält.
     


    Revendications

    1. Support non transitoire lisible par ordinateur comprenant des instructions, qui lorsque exécutées par un processeur, mettent en oeuvre un procédé, le procédé comprenant :

    créer (402) un objet de transaction pour une transaction à effectuer pour un objet,
    dans lequel l'objet de transaction est identifié par une ID objet de transaction (TOI) et associé à l'objet identifié par une ID objet (OID) ;

    stocker (408), dans un stockage rémanent, une première entrée de table des matières (TE) et un fragment de métadonnées (MD) pour l'objet de transaction, où la première TE fait référence au fragment MD, où le fragment MD correspond à une quantité finie de métadonnées et identifie l'objet ;

    recevoir (412) une requête d'écriture pour écrire des données à l'objet de transaction comme faisant partie de la transaction, où la requête d'écriture spécifie la TOI et un décalage ;

    en réponse à la requête d'écriture :

    stocker (414, 416), dans le stockage rémanent, une deuxième TE et un fragment de données, où la deuxième TE spécifie la TOI et le décalage, et où la deuxième TE fait référence au fragment de données, où le fragment de données correspond à une quantité finie de données utilisateur et comprend les données ;

    stocker (414, 416), dans une structure de données en mémoire située dans une mémoire connectée à un contrôleur de stockage, une entrée comprenant une valeur de hachage et une adresse physique du fragment de données, où la valeur de hachage est dérivée de la TOI et du décalage ;

    recevoir (422) une requête d'engagement pour engager la transaction ;

    en réponse à la requête d'engagement :

    stocker (428), dans un stockage rémanent, une troisième TE et un deuxième fragment de métadonnées (MD) pour l'objet de transaction, où la troisième TE fait référence au deuxième fragment MD, où le deuxième fragment MD identifie l'objet et spécifie que la transaction est engagée ;

    actualiser (430), dans la structure de données en mémoire, une deuxième entrée comprenant une deuxième valeur de hachage et une deuxième adresse physique pour un deuxième fragment de données pour remplacer la deuxième adresse physique par l'adresse physique, où la deuxième valeur de hachage est dérivée de l'OID et du décalage ;

    dans lequel avant que la requête d'engagement ne soit traitée, des données dans l'objet sont accessibles en utilisant l'OID et le décalage et les données dans le fragment de données associé à la transaction, mais non engagée, sont accessibles en utilisant la TOI et le décalage ; et

    dans lequel après que la requête d'engagement a été traitée, les données dans le fragment de données sont accessibles en utilisant l'OID et le décalage et des deuxièmes données dans le deuxième fragment de données ne sont pas accessibles.


     
    2. Support non transitoire lisible par ordinateur selon la revendication 1, dans lequel le stockage rémanent est une mémoire transistorisée.
     
    3. Support non transitoire lisible par ordinateur selon la revendication 1, dans lequel la requête d'engagement spécifie la TOI et ne spécifie pas l'OID.
     
    4. Support non transitoire lisible par ordinateur selon la revendication 1, dans lequel stocker la première TE comprend stocker la première TE dans une page de table des matières (TOC) dans le stockage rémanent.
     
    5. Support non transitoire lisible par ordinateur selon la revendication 1, dans lequel la première TE et le fragment MD sont stockés dans des pages séparées dans un bloc dans le stockage rémanent.
     
    6. Support non transitoire lisible par ordinateur selon la revendication 1, dans lequel le fragment MD identifie l'objet en utilisant l'OID.
     
    7. Support non transitoire lisible par ordinateur selon la revendication 1, dans lequel avant de traiter la requête d'engagement, le fragment MD spécifie que la transaction n'est pas engagée.
     
    8. Support non transitoire lisible par ordinateur selon la revendication 1, dans lequel la première TE comprend un troisième décalage égal à -1.
     




    Drawing

























































    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description