WO2006044706A2

WO2006044706A2 - System and method for reclaming unused space from a thinly provisioned data container

Info

Publication number: WO2006044706A2
Application number: PCT/US2005/037093
Authority: WO
Inventors: Vijayan Rajan; David Brittain Bolen; Douglas P. Doucette
Original assignee: Network Appliance, Inc.
Priority date: 2004-10-15
Filing date: 2005-10-13
Publication date: 2006-04-27
Also published as: US8621172B2; WO2006044706A3; US7603532B2; US20090248763A1; JP5054531B2; US20060085471A1; JP2008517383A; EP1805591A2

Abstract

A system and method for reclaims unused space from a thinly provision data container served by a storage system. A host-side agent detects blocks of the data container that may be freed and sends a novel Punch Hole command to the storage system associated with the data container. The storage system allocates the appropriate blocks in response to the Punch Hole command.

Description

SYSTEM AND METHOD FOR RECLAIMING UNUSED SPACE FROM A THINLY PROVISIONED DATA CONTAINER

FIELD OF THE INVENTION

The present invention relates to storage systems and, in particular, to reclaiming 5 unused space from a thinly provisioned data container on a storage system.

BACKGROUND OF THE INVENTION

A storage system is a computer that provides storage service relating to the or¬ ganization of information on writable persistent storage devices, such as memories, tapes or disks. The storage system is commonly deployed within a storage area network (SAN)

I₀ or a network attached storage (NAS) environment. When used within a NAS environ¬ ment, the storage system may be embodied as a file server including an operating system that implements a file system to logically organize the information as a hierarchical struc¬ ture of directories and files on, e.g. the disks. Each "on-disk" file may be implemented as a set of data structures, e.g., disk blocks, configured to store information, such as the ac-

I₅ tual data for the file. A directory, on the other hand, may be implemented as a specially formatted file in which information about other files and directories are stored. As used herein a file is defined to be any logical storage container that contains a fixed or variable amount of data storage space, and that may be allocated storage out of a larger pool of available data storage space. As such, the term file, as used herein and unless the context

20 otherwise dictates, can also mean a container, object or any other storage entity that does not correspond directly to a set of fixed data storage devices. A file system is, generally, a computer system for managing such files, including the allocation of fixed storage space to store files on a temporary or permanent basis.

25 The storage system may be further configured to operate according to a cli¬ ent/server model of information delivery to thereby allow many client systems (clients) to access shared resources, such as files, stored on the storage system. Sharing of files is a hallmark of a NAS system, which is enabled because of its semantic level of access to files and file systems. Storage of information on a NAS system is typically deployed over a computer network comprising a geographically distributed collection of intercon¬ nected communication links, such as Ethernet, that allow clients to remotely access the information (files) on the filer. The clients typically communicate with the storage sys- tern by exchanging discrete frames or packets of data according to pre-defined protocols, such as the Transmission Control Protocol/Internet Protocol (TCP/IP).

In the client/server model, the client may comprise an application executing on a computer that "connects" to the storage system over a computer network, such as a point- to-point link, shared local area network, wide area network or virtual private network im- plemented over a public network, such as the Internet. NAS systems generally utilize file-based access protocols; therefore, each client may request the services of the storage system by issuing file system protocol messages (in the form of packets) to the file sys¬ tem over the network identifying one or more files to be accessed without regard to spe¬ cific locations, e.g., blocks, in which the data are stored on disk. By supporting a plural- ity of file system protocols, such as the conventional Common Internet File System

(CIFS), the Network File System (NFS) and the Direct Access File System PAFS) pro¬ tocols, the utility of the storage system may be enhanced for networking clients.

A SAN is a high-speed network that enables establishment of direct connections between a storage system and its storage devices. The SAN may thus be viewed as an extension to a storage bus and, as such, an operating system of the storage system enables access to stored information using block-based access protocols over the "extended bus". In this context, the extended bus is typically embodied as Fibre Channel (FC) or Ethernet media adapted to operate with block access protocols, such as Small Computer Systems Interface (SCSI) protocol encapsulation over FC or TCP/IP/Ethernet. A SAN arrangement or deployment allows decoupling of storage from the storage system, such as an application server, and some level of information storage sharing at the application server level. There are, however, environments wherein a SAN is dedi¬ cated to a single server. In some SAN deployments, the information is organized in the form of databases, while in others a file-based organization is employed. Where the in- formation is organized as files, the client requesting the information maintains file map- pings and manages file semantics, while its requests (and server responses) address the information in terms of block addressing on disk using, e.g., a logical unit number (LUN).

In some SAN environments, storage systems may export virtual disks (vdisks) to clients utilizing block-based protocols, such as, for example, Fibre Channel and iSCSI. One example of a vdisk is a special file type in a volume that derives from a plain file, but that has associated export controls and operation restrictions that support emulation of a disk. Vdisks are described further in U.S. Patent Application Serial No. 10/216,453, entitled STORAGE VIRTUALIZATION BY LAYERING VIRTUAL DISK OBJECTS ON A FILE SYSTEM, by Vijayan Rajan, et al., the contents of which are hereby incor¬ porated by reference. These block-based protocols and the exported file/vdisks appear as physical disk devices to the clients of the storage system.

Certain file systems, including the exemplary write anywhere file layout (WAFL) file system available from Network Appliance, Inc, of Sunnyvale, CA, include the capa- bility to generate a thinly provisioned data container, wherein the data container is not completely written to disk at the time of its creation. As used herein, the term data con¬ tainer generally refers to a unit of storage for holding data, such as a file system, disk file, volume or a logical number (LUN), which is addressable by, e.g., its own unique identifi¬ cation. The storage space required to hold the data contents of the thinly provisioned data container on disk has not yet been used. The use of thinly provisioned data container is often utilized in the exemplary WAFL file system environment when, for example, a vdisk is initially generated. A user or administrator may generate a vdisk of specified size, for example, 10 gigabytes (GB). This size represents the maximum addressable space of the vdisk. To increase system performance, the file system generally does not write the entire vdisk to the disks at the time of creation. Instead, the file system gener¬ ates a thinly provisioned data container (i.e., file) representing the vdisk. The thinly pro¬ visioned data container may then be populated (filled in) via subsequent write operations as the vdisk is filled in with data. While this description is written in terms of a thinly provisioned data container over and underlying file system, it should be noted that other thin provisioning implementations may be utilized. As such, the use of an underlying file system to support a thinly provisioned data container should be taken as exemplary only. Fig. 1 is a schematic block diagram of an (inode structure) buffer tree 100 of an exemplary thinly provisioned data container. This (inode) buffer tree structure 100 is cre¬ ated when, for example, a vdisk is first created by the file system as thinly provisioned. In a typical thinly provisioned data container, only the inode 105 is actually written to disk. The remainder of the data container is not written to or otherwise physically stored on the disks storing the data container. The data container 100 includes a completed inode 105, however, it does not contain indirect blocks 110, 120 or file data blocks 125 (as shown in phantom). Thus, these phantom blocks (i.e., 110, 120, 125) are not gener¬ ated when the data container is created, although, they will be written to disk as the data container is populated. By only writing the inode to disk when a thinly provisioned data container is generated, substantial time is saved as the number of disk accesses is re¬ duced. Additionally, only the storage space on the disks that is needed to hold the con¬ tents of the data container are utilized. Illustratively, the file system will make appropri¬ ate space reservations to ensure that the entire thinly provisioned data container may be written to disk. Space reservation techniques are described in U.S. Patent Application Serial No. 10/423,391, entitled SYSTEM AND METHOD FOR RESERVING SPACE TO GUARANTEE FILE WRIT ABILITY IN A FILE SYSTEM SUPPORTING PERSISTENT CONSISTENCY POINT IMAGES, by Peter F. Corbett, et al.

Fig. 2 is a schematic block diagram of an exemplary (inode) buffer tree structure 200 of a partially filled in thinly provisioned data container that includes original inode 105. Here, indirect blocks 210, 220 and exemplary file data block 225 have been popu¬ lated (filled in) in response to one or more write operations to the data container. Contin¬ ued write operations will result in filling in additional data blocks, for example, file data block 325 as shown in the exemplary (inode) buffer tree structure 300 of Fig. 3. Eventu- ally, when the data container has been completely filled, all blocks, including such blocks as indirect blocks 420 and associated file data blocks (not shown) will be completed as illustrated in the schematic block diagram of an exemplary inode structure 400 in Fig. 4. At such time, the thinly provisioned data container has been completely filled in and each block is associated with an actual block on disk. A known environment for utilizing a storage system with a thinly provisioned data container, i.e., a thinly provisioned LUN, involves overlaying a host-side file system onto the thinly provisioned LUN. In such an environment, the host (or client of the stor¬ age system) includes a file system that utilizes the exported LUN as storage and main¬ tains structured storage, e.g., a file system, on the blocks of the LUN. However, a noted disadvantage is that the host-side file system does not communicate status to the storage system concerning the deletion or deallocation of blocks within the LUN. Although the file system typically records appropriate metadata entries when a file is deleted, no status message is passed to the storage system that notifies the system that certain blocks of the LUN are no longer in use. Thus, while the LUN may dynamically grow by allocating additional blocks (up to its maximum number of addressable blocks) as needed, it will not deallocate blocks as files are deleted in the host-side file system. For example, if a LUN is generated with a maximum size of 100GB and then a 50GB file is written to it, the LUN will allocate 50GB of space on the storage system. If the 50GB file is thereafter deleted in the host-side file system, that file system records appropriate metadata entries and frees its file system pointers. However, the LUN will still occupy 50GB of space on the storage system, even though the 50GB is now unused space within the LUN.

SUMMARY OF THE INVENTION

The disadvantages of the prior art are overcome by providing a system and method for reclaiming unused storage space from a thinly provisioned data container, such as a logical unit number (LUN) of a storage system. A host-side agent executes on a client (host) of the storage system. The host-side agent detects which blocks have been freed from a host-side file system and sends a novel Punch Hole command to the storage system, which causes the storage system to deallocate certain ranges of blocks within the data container, thereby permitting the data container to shrink in size. The agent sends the Punch Hole command to the storage system via a conventional data pathway between the client and the storage system, e.g., as a vendor-specific SCSI command over a FCP connection.

In an alternate embodiment, the agent iteratively allocates a file on the host-side file system, locks the file and determines which blocks of the underlying data container on the storage system are supporting the locked file. The agent then sends the novel Punch Hole command to the storage system to deallocate the blocks associated with the locked file. By repeatedly performing this process and ensuring that the files are stored on differing blocks of the data container, the agent may ensure that all unused blocks of data container are freed.

Additionally, the agent may interface with a host-side application that does not implement a file system but utilizes some other form of structured storage, such as a da¬ tabase program. In such an embodiment, the agent queries the application to determine the nature of the structured storage utilized by the application and then sends one or more appropriate Punch Hole commands to the storage system to deallocate any unused blocks of the data container.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and further advantages of the invention may be better understood by referring to the following description in conjunction with the accompanying drawings in which like reference numerals indicate identical or functionally similar elements: Fig.1 , already described, is a schematic block diagram of an exemplary thinly provisioned data container showing a inode for the data container;

Fig. 2, already described, is a schematic block diagram of a partially filled in thinly provisioned data container in accordance with an embodiment of the present inven¬ tion; Fig. 3 is a schematic block diagram of a an exemplary partially filled in thinly provisioned data container in accordance with an embodiment of the present invention;

Fig. 4, already described, is a schematic block diagram of an exemplary filled in data container in accordance with an embodiment of the present invention;

Fig. 5 is a schematic block diagram of an exemplary storage system in accor- dance with an embodiment of the present invention;

Fig. 6 is a schematic block diagram of an exemplary storage operating system for use with the storage system of Fig.5 in accordance with an embodiment of the present invention; Fig. 7 A is a schematic block diagram of the format of an exemplary Punch Hole command structure in accordance with an embodiment of the present invention;

Fig. 7B is a schematic block diagram of the format of an exemplary Punch Hole command structure in accordance with an embodiment of the present invention; Fig. 8 is a flowchart detailing the steps of a procedure for reclaiming unused space in a thinly provisioned data container in accordance with an embodiment of the present invention; and

Fig. 9 is a flowchart detailing the steps of a procedure for reclaiming unused space in a thinly provisioned data container in accordance with an embodiment of the present invention.

DETAILED DESCRIPTION OF AN ILLUSTRATIVE EMBODIMENT

A. Storage Appliance

The present invention may be implemented, in the illustrative embodiment, on a storage appliance that serves both file and block protocol access to information stored on storage devices in an integrated manner. In this context, the term storage appliance de¬ notes a computer having features such as simplicity of storage service management and ease of storage reconfiguration, including reusable storage space, for users (system ad¬ ministrators) and clients of network attached storage (NAS) and storage area network (SAN) deployments. The storage appliance may provide NAS services through a file system, while the same appliance provides SAN services through SAN virtualization, in¬ cluding logical unit number (LUN) emulation. While this description is written in terms of storage appliances, the principles of the present invention may be applied to any stor¬ age system. As such the use of storage appliances should be taken as exemplary only.

Fig. 5 is a schematic block diagram of a storage appliance 500 configured to pro- vide storage service relating to the organization of information on storage devices, such as disks 530. The storage appliance 500 is illustratively embodied as a storage system comprising a processor 522, a memory 524, a plurality of network adapters 525, 526 and a storage adapter 528 interconnected by a system bus 523. The multi-protocol storage appliance 500 also includes a storage operating system 600 that provides a virtualization system (and, in particular, a file system) to logically organize the information as a hierar¬ chical structure of named directory, file and virtual disk (vdisk) storage objects on the disks 530.

5 Whereas clients of a NAS-based network environment have a storage viewpoint of files, the clients of a SAN-based network environment have a storage viewpoint of blocks or disks. To that end, the storage appliance 500 presents (exports) disks to SAN clients through the creation of logical unit numbers (LUNs) or vdisk objects. A vdisk object (hereinafter "vdisk") is a special file type that is implemented by the virtualization

I₀ system and translated into an emulated disk as viewed by the SAN clients. The storage appliance thereafter makes these vdisks accessible to the SAN clients through controlled exports, as described further herein.

In the illustrative embodiment, the memory 524 comprises storage locations that are addressable by the processor and adapters for storing software program code and data is structures associated with the present invention. The processor and adapters may, in turn, comprise processing elements and/or logic circuitry configured to execute the software code and manipulate the data structures. The storage operating system 600, portions of which are typically resident in memory and executed by the processing elements, func¬ tionally organizes the storage appliance by, inter alia, invoking storage operations in

20 support of the storage service implemented by the appliance. It will be apparent to those skilled in the art that other processing and memory means, including various computer readable media, may be used for storing and executing program instructions pertaining to the inventive system and method described herein.

The network adapter 525 couples the storage appliance to a plurality of clients 2S 560a,b over point-to-point links, wide area networks, virtual private networks imple¬ mented over a public network (Internet) or a shared local area network, hereinafter re¬ ferred to as an illustrative Ethernet network 565. Therefore, the network adapter 525 may comprise a network interface card (NIC) having the mechanical, electrical and signaling circuitry needed to connect the appliance to a network switch, such as a conventional 30 Ethernet switch 570. For this NAS-based network environment, the clients are config- ured to access information stored on the multi-protocol appliance as files. The clients 560 communicate with the storage appliance over network 565 by exchanging discrete frames or packets of data according to pre-defined protocols, such as the Transmission Control Protocol/Internet Protocol (TCP/IP). The clients 560 may be general-purpose computers configured to execute applica¬ tions over a variety of operating systems, including the UNIX® and Microsoft® Win¬ dows™ operating systems. Client systems generally utilize file-based access protocols when accessing information (in the form of files and directories) over a NAS-based net¬ work. Therefore, each client 560 may request the services of the storage appliance 500 by issuing file access protocol messages (in the form of packets) to the appliance over the network 565. It will be apparent to those skilled in the art that other clients running other types of operating systems may also communicate with the integrated multi-protocol storage appliance using other file access protocols.

Illustratively, client (or host) 560b includes a file system 590 that interfaces with one or more applications 592. The host-side file system 590 illustratively implements a file system overlaid onto a data container serviced by the storage system. For example, the storage system may export a LUN, which the host-side file system 590 utilizes to store data. In an illustrative embodiment, a novel host-side agent 594 also executes on client 560b. According to the invention, the agent 594 blocks of a thinly provisioned data container may be reclaimed and by sending a novel Punch Hole command to the storage system, as described further below. Alternately, a non-file system application 596 exe¬ cuting on client 560a, which application 596 may comprise a database system or other system. In accordance with an alternate embodiment of the present invention, the novel agent 594 may also execute on client 560a The storage network "target" adapter 526 also couples the multi-protocol storage appliance 500 to clients 560 that may be further configured to access the stored informa¬ tion as blocks or disks. For this SAN-based network environment, the storage appliance is coupled to an illustrative Fibre Channel (FC) network 585. FC is a networking stan¬ dard describing a suite of protocols and media that is primarily found in SAN deploy- ments. The network target adapter 526 may comprise a FC host bus adapter (HBA) hav- ing the mechanical, electrical and signaling circuitry needed to connect the appliance 100 to a SAN network switch, such as a conventional FC switch 580. In addition to providing FC access, the FC HBA may offload Fibre Channel network processing operations for the storage appliance. The clients 560 generally utilize block-based access protocols, such as the Small

Computer Systems Interface (SCSI) protocol, when accessing information (in the form of blocks, disks or vdisks) over a SAN-based network. SCSI is a peripheral input/output (I/O) interface with a standard, device independent protocol that allows different periph¬ eral devices, such as disks 530, to attach to the storage appliance 500. In SCSI terminol- ogy, clients 560 operating in a SAN environment are initiators that initiate requests and commands for data. The multi-protocol storage appliance is thus a target configured to respond to the requests issued by the initiators in accordance with a request/response pro¬ tocol. The initiators and targets have endpoint addresses that, in accordance with the FC protocol, comprise worldwide names (WWN). A WWN is a unique identifier, e.g., a node name or a port name, consisting of an 8-byte number.

The storage appliance 500 supports various SCSI-based protocols used in SAN deployments, including SCSI encapsulated over TCP (iSCSI) and SCSI encapsulated over FC (FCP). The initiators (hereinafter clients 560) may thus request the services of the target (hereinafter storage appliance 500) by issuing iSCSI and FCP messages over the network 565, 585 to access information stored on the disks. It will be apparent to those skilled in the art that the clients may also request the services of the integrated multi-protocol storage appliance using other block access protocols. By supporting a plu¬ rality of block access protocols, the multi-protocol storage appliance provides a unified and coherent access solution to vdisks/LUNs in a heterogeneous SAN environment. The storage adapter 528 cooperates with the storage operating system 600 execut¬ ing on the storage appliance to access information requested by the clients. The informa¬ tion may be stored on the disks 530 or other similar media adapted to store information. The storage adapter includes I/O interface circuitry that couples to the disks over an I/O interconnect arrangement, such as a conventional high-performance, FC serial link topol- ogy. The information is retrieved by the storage adapter and, if necessary, processed by the processor 522 (or the adapter 528 itself) prior to being forwarded over the system bus 523 to the network adapters 525, 526, where the information is formatted into packets or messages and returned to the clients.

Storage of information on the appliance 500 is preferably implemented as one or more storage volumes (e.g., VOL1-2 550) that comprise a cluster of physical storage disks 530, defining an overall logical arrangement of disk space. The disks within a vol¬ ume are typically organized as one or more groups of Redundant Array of Independent (or Inexpensive) Disks (RAID). RAID implementations enhance the reliability/integrity of data storage through the writing of data "stripes" across a given number of physical disks in the RAID group, and the appropriate storing of redundant information with re¬ spect to the striped data. The redundant information enables recovery of data lost when a storage device fails. It will be apparent to those skilled in the art that other redundancy techniques, such as mirroring, may be used in accordance with the present invention.

Specifically, each volume 550 is constructed from an array of physical disks 530 that are organized as RAID groups 540, 542, and 544. The physical disks of each RAID group include those disks configured to store striped data (D) and those configured to store parity (P) for the data, in accordance with an illustrative RAID 4 level configura¬ tion. It should be noted that other RAID level configurations (e.g. RAID 5) are also con¬ templated for use with the teachings described herein. In the illustrative embodiment, a minimum of one parity disk and one data disk may be employed.

B. Storage Operating System

To facilitate access to the disks 530, the storage operating system 600 implements a write-anywhere file system of a virtualization system that "virtualizes" the storage space provided by disks 530. The file system logically organizes the information as a hierarchical structure of named directory and file objects (hereinafter "directories" and "files") on the disks. Each "on-disk" file may be implemented as set of disk blocks con¬ figured to store information, such as data, whereas the directory may be implemented as a specially formatted file in which names and links to other files and directories are stored. The virtualization system allows the file system to further logically organize information as a hierarchical structure of named vdisks on the disks, thereby providing an integrated NAS and SAN appliance approach to storage by enabling file-based (NAS) access to the named files and directories, while further enabling block-based (SAN) access to the named vdisks on a file-based storage platform. The file system simplifies the complexity of management of the underlying physical storage in SAN deployments.

As noted, a vdisk is a special file type in a volume that derives from a plain (regu¬ lar) file, but that has associated export controls and operation restrictions that support emulation of a disk. Unlike a file that can be created by a client using, e.g., the NFS or CIFS protocol, a vdisk is created on the storage appliance via, e.g. a user interface (UI) as a special typed file (object). Illustratively, the vdisk is a multi-inode object comprising a special file inode that holds data and at least one associated stream inode that holds at¬ tributes, including security information. The special file inode functions as a main con¬ tainer for storing data, such as application data, associated with the emulated disk. The stream inode stores attributes that allow LUNs and exports to persist over, e.g., reboot operations, while also enabling management of the vdisk as a single disk object in rela¬ tion to SAN clients. An example of a vdisk and its associated inodes that may be advan¬ tageously used with the present invention is described in U.S. Patent Application Serial No. 10/216,453, entitled STORAGE VIRTUALIZATION BY LAYERING VDISKS ON A FILE SYSTEM, by which application is hereby incorporated by reference as though fully set forth herein.

In accordance with an illustrative embodiment of the present invention, when a vdisk is generated it is typically created as a thinly provisioned data container. However, the storage operating system will also reserve the appropriate amount of storage space to fill the "holes" of the newly generated vdisk. This space reservation technique ensures that there is sufficient space on the disks to completely fill in the data container. Exem¬ plary space reservation policies and techniques are further described in U.S. Patent Ap¬ plication Serial No. 10/423,391, entitled SYSTEM AND METHOD FOR RESERVING SPACE TO GURANTEE FILE WRITABILITY IN A FILE SYSTEM SUPPORTING PERSISITENT CONSISTENCY POINT IMAGES, by Peter F. Corbett, et al. In the illustrative embodiment, the storage operating system is preferably the NetApp® Data ONT AP™ operating system available from Network Appliance, Inc., Sunnyvale, California that implements a Write Anywhere File Layout (WAFL™) file sys¬ tem. However, it is expressly contemplated that any appropriate storage operating sys- tern, including a write in-place file system, may be enhanced for use in accordance with the inventive principles described herein. As such, where the term "WAFL" is employed, it should be taken broadly to refer to any file system that is otherwise adaptable to the teachings of this invention.

As used herein, the term "storage operating system" generally refers to the com- puter-executable code operable on a computer that manages data access and may, in the case of a multi-protocol storage appliance, implement data access semantics, such as the Data ONTAP storage operating system, which is implemented as a microkernel. The storage operating system can also be implemented as an application program operating over a general-purpose operating system, such as UNIX® or Windows NT®, or as a gen- eral-purpose operating system with configurable functionality, which is configured for storage applications as described herein.

In addition, it will be understood to those skilled in the art that the inventive sys¬ tem and method described herein may apply to any type of special-purpose (e.g., storage serving appliance) or general-purpose computer, including a standalone computer or por- tion thereof, embodied as or including a storage system. Moreover, the teachings of this invention can be adapted to a variety of storage system architectures including, but not limited to, a network-attached storage environment, a storage area network and disk as¬ sembly directly-attached to a client or host computer. The term "storage system" should therefore be taken broadly to include such arrangements in addition to any subsystems configured to perform a storage function and associated with other equipment or systems.

Fig. 6 is a schematic block diagram of an exemplary storage operating system 600 that may be advantageously used with the present invention. The storage operating sys¬ tem comprises a series of software layers organized to form an integrated network proto¬ col stack or, more generally, a multi-protocol engine that provides data paths for clients to access information stored on the multi-protocol storage appliance using block and file access protocols. The protocol stack includes a media access layer 610 of network driv¬ ers (e.g., gigabit Ethernet drivers) that interfaces to network protocol layers, such as the IP layer 612 and its supporting transport mechanisms, the TCP layer 614 and the User Datagram Protocol (UDP) layer 616. A file system protocol layer provides multi- protocol file access and, to that end, includes support for the DAFS protocol 618, the NFS protocol 620, the CIFS protocol 622 and the Hypertext Transfer Protocol (HTTP) protocol 624. A Virtual Interface (VI) layer 626 implements the VI architecture to pro¬ vide direct access transport (DAT) capabilities, such as RDMA, as required by the DAFS protocol 618. An iSCSI driver layer 628 provides block protocol access over the TCP/IP net¬ work protocol layers, while a FC driver layer 630 operates with the FC HBA 526 to re¬ ceive and transmit block access requests and responses to and from the integrated storage appliance. The FC and iSCSI drivers provide FC-specific and iSCSI-specific access con¬ trol to the LUNs (vdisks) and, thus, manage exports of vdisks to either iSCSI or FCP or, alternatively, to both iSCSI and FCP when accessing a single vdisk on the multi-protocol storage appliance. In addition, the storage operating system includes a disk storage layer 640 that implements a disk storage protocol, such as a RAID protocol, and a disk driver layer 650 that implements a disk access protocol such as, e.g., a SCSI protocol.

Bridging the disk software layers with the integrated network protocol stack lay- ers is a virtualization system 655 that is implemented by a file system 665 cooperating with virtualization modules illustratively embodied as, e.g., vdisk module 670 and SCSI target module 660. It should be noted that the vdisk module 670, file system 665 and SCSI target module 660 can be implemented in software, hardware, firmware, or a com¬ bination thereof. The vdisk module 670 is layered on (and interacts with) the file system 665 to provide a data path from the block-based SCSI target module to blocks managed by the file system. In essence, the vdisk module 670 manages SAN deployments by, among other things, implementing a comprehensive set of vdisk (LUN) commands issued through a user interface by a system administrator. These vdisk commands are converted to primitive file system operations ("primitives") that interact with the file system 665 and the SCSI target module 660 to implement the vdisks. The SCSI target module 660, in turn, initiates emulation of a disk or LUN by pro¬ viding a mapping procedure that translates logical block access to LUNs specified in ac¬ cess requests into virtual block access to the special vdisk file types and, for responses to the requests, vdisks into LUNs. The SCSI target module is illustratively disposed be- tween the FC and iSCSI drivers 630, 628 and the file system 665 to thereby provide a translation layer of the virtualization system 655 between the SAN block (LUN) space and the file system space, where LUNs are represented as vdisks. Additionally, in the illustrative embodiment, the SCSI target module 660 interprets a novel received Punch Hole command from a host-side agent and implements deallocation of blocks, in con- junction with the file system 665 and vdisk module 670, that are no longer in use. As de¬ scribed further below, the novel Punch Hole command permits a thinly provisioned data container to reduce the number of blocks currently allocated by it as the amount of struc¬ tured storage overlaid onto the data container decreases.

The file system 665 illustratively implements the above-described WAFL file sys- tern having an on-disk format representation that is block-based using, e.g., 4 kilobyte (kB) blocks and using inodes to describe the files. Broadly stated, all inodes of the file system are organized into the inode file. A file system (fs) info block specifies the layout of information in the file system and includes an inode of a file that includes all other inodes of the file system. Each volume has an fsinfo block that is preferably stored at a fixed location within, e.g., a RAID group of the file system. The inode of the root fsinfo block may directly reference (point to) blocks of the inode file or may reference indirect blocks of the inode file that, in turn, reference direct blocks of the inode file. Within each direct block of the inode file are embedded inodes, each of which may reference indirect blocks that, in turn, reference data blocks of a file or vdisk. It should be noted that the software "path" through the storage operating system layers described above needed to perform data storage access for the client request re¬ ceived at the multi-protocol storage appliance may alternatively be implemented in hard¬ ware. That is, in an alternate embodiment of the invention, a storage access request data path through the operating system layers (including the virtualization system 655) may be implemented as logic circuitry embodied within a field programmable gate array (FPGA) or an application specific integrated circuit (ASIC). This type of hardware implementa- tion increases the performance of the storage service provided by appliance 500 in re¬ sponse to a file access or block access request issued by a client 560. Moreover, in an¬ other alternate embodiment of the invention, the processing elements of network and storage adapters 525-528 may be configured to offload some or all of the packet process- ing and storage access operations, respectively, from processor 522 to thereby increase the performance of the storage service provided by the multi-protocol storage appliance. It is expressly contemplated that the various processes, architectures and procedures de¬ scribed herein can be implemented in hardware, firmware or software.

C. Reclaiming Unused Space From A Thinly Provisioned Data Container The present invention is directed to a system and method for reclaiming unused storage space from a thinly provisioned data container. The invention enables a thinly provisioned data container stored on a storage system to reduce the number of blocks al¬ located to it as the structured storage e.g., a host side file system overlaid onto the data container decreases. In an illustrative embodiment, a host-side agent executes on a client of the storage system and determines appropriate blocks that may be reclaimed due to the overlaid structured storage no longer utilizing them. The agent then generates the novel Punch Hole command and sends it to the storage system using the conventional data pathway between the client and the storage system. Illustratively, the Punch Hole com¬ mand is implemented as a vendor-specific SCSI command but, in alternate embodiments, may be implemented using other techniques. For clients utilizing a non-file system ap¬ plication, such as a database application, the host-side agent interfaces with the applica¬ tion to determine appropriate blocks of the data container that may be reclaimed due to the host-side application no longer utilizing the storage space. The agent then generates and sends the appropriate Punch Hole command to the storage system. Fig. 7 A is a schematic block diagram of the format of the novel Punch Hole com¬ mand structure 700A in accordance with an embodiment of the present invention. As noted, the Punch Hole command is illustratively implemented as a vendor specific SCSI command. However, it should be noted that the Punch Hole command may be imple¬ mented in other ways including, for example, implementing the command into another protocol specification. The Punch Hole command structure 700A includes an operation code field 705, a number of a ranges field 710, a control field 715, a first range field 720 that includes a logical block address field 725 and a range length field 730 and, in alter¬ nate embodiments, additional field 735. The operation code field 705 identifies the command as a Punch Hole command. The number of ranges field 710 identifies the number of range values included in this command. The control field 715 is utilized to pass control information to the storage system. The first range field 720 includes two sub fields, namely, a logical block address field 725 and a range length field 730. The logical block address field 725 identifies the starting point of the range of blocks to be allocated, whereas the range length field 730 identifies the number of blocks to be deallocated. In the illustrative embodiment, there are as many range fields, such as first range field 720, as there are identified ranges in the number of ranges field 710. The Punch Hole com¬ mand 700A identifies one or more ranges of blocks, each starting at the logical block ad¬ dress identified in field 725 and continuing on for the number of blocks identified in field 730. Similarly, Fig. 7B is a schematic block diagram of the format of the Punch Hole command structure 700B in accordance with an alternate embodiment of the present in¬ vention. The Punch Hole command structure 700B includes an operation code field 705, a bitmap size field 740, a control field 715, a logical block address field 745, a bitmap field 750 and, in alternate embodiments, additional field 735. The operation code field 705 identifies the command as a Punch Hole command. The bitmap size field 740 identi¬ fies the size of the bitmap contained in bitmap field 750. The control field 715 is utilized to pass control information to the storage system. The logical block address field 745 identifies a starting block address. The bitmap field 750 contains a bitmap wherein each bit represents a single block. In this alternate embodiment, the storage system deter- mines which block to deallocate by adding the offset of the bit in the bitmap to the logical block address contained in the command. If the bit is set, the corresponding block is de¬ allocated. As noted above, these illustrative command structures 700 A,B are exemplary only. The novel Punch Hole command may be implemented using other structures as will be appreciated by one skilled in the art. Fig. 8 is a flowchart detailing the steps of a procedure 800 for reclaiming unused space from a thinly provisioned data container in accordance with an embodiment of the present invention. The data container is illustratively described herein as a logical unit number (LUN); however, it should be noted that any suitable data container may be util¬ ized in accordance with the principles of the present invention. As such, the term LUN should not be taken to be limiting and any suitable data container may be utilized. The procedure 800 begins in step 805 and continues to step 810 where the agent executing on a client of the storage system determines that blocks are no longer in use on the LUN. This may occur by, for example, the host-side agent querying the file system or by exam¬ ining file system metadata to determine blocks that have been freed within the file sys¬ tem. Alternately, in embodiments where the host-side agent is executing on a client util- izing a non-file system application, the agent may determine that blocks are no longer in use on the LUN by querying the application and/or analyzing its structured storage meta¬ data.

Upon identifying a number of blocks that are no longer in use on the LUNs, the agent generates and sends a novel Punch Hole command directed to the LUN (step 815). The generated Punch Hole command identifies the appropriate ranges of blocks to be freed. The Punch Hole command is typically sent via the conventional data pathway be¬ tween the client and the storage system. For example, if the client normally communi¬ cates using FCP with the storage system, the agent will generate a Punch Hole command and send it using the FCP protocol. In response, the storage system releases the identi- fied ranges of a blocks in the LUN and deallocates the underlying blocks to be reused by the storage system. These blocks are typically deallocated by updating appropriate file system metadata to show that they may be re-used by the file system. Additionally, any pointers to the blocks, such as pointers in high level indirect blocks are cleared. In step 825, the storage system replies with a response message either acknowledging that the command was successful or with an appropriate error code. The procedure then com¬ pletes in step 830.

In an alternate embodiment, the agent is more proactive regarding the identifica¬ tion of blocks that may be freed. Fig. 9 is a flowchart detailing the steps of a procedure 900 for reclaiming unused space from a thinly provisioned data container in accordance with an alternate embodiment of the invention. The procedure 900 begins in step 905 continues to step 910 where the agent allocates a file on the host-side file system. This may be accomplished using conventional file system operation commands to generate a file of a predetermined size. In step 915, the agent locks the allocated file so that no op¬ erations may be directed to it. This lock may be generated using conventional host-side file system commands. In step 920, the agent determines which data container blocks store the allocated file, by, e.g., using conventional file system operations to determine the location within the file system of a particular file. In step 925, the agent prepares and sends a Punch Hole command to the storage system identifying the ranges of blocks in which the allocated file resides. The agent then frees the allocated file in step 930 before the procedure completes in step 935. By freeing the allocated file, the host-side file sys- tern updates the appropriate pointers to indicate that the blocks previously utilized by the file are no longer in use. Similarly, as result of the Punch Hole command being proc¬ essed by the storage system, the underlying blocks of the data container of the storage system are also freed and may be reutilized by the storage system.

To again summarize, the present invention provides a system and method for re- claiming unused space in a thinly provisioned data container on a storage system. A host-side agent determines blocks of the structured storage of the client that may be de¬ allocated on the data container by, e.g., querying the host-side file system or by creating a file and determining the blocks storing the created file. The agent then generates a novel Punch Hole command identifying the blocks to be de-allocated on the data container and sends the Punch Hole command to the storage system serving the data container. In re¬ sponse to receiving the Punch Hole command, the storage system deallocates the identi¬ fied blocks (or ranges of blocks) on the data container so that the data container con¬ sumes less storage space, thereby enabling the container to dynamically grow and shrink in accordance with the amount of data being stored thereon. The foregoing description has been directed to specific embodiments of this in¬ vention. It will be apparent, however, that other variations and modifications may be made to the described embodiments, with the attainment of some or all of their advan¬ tages. For example, it is expressly contemplated that the teachings of this invention can be implemented as software, including a computer-readable medium having program in- structions executing on a computer, hardware, firmware, or a combination thereof. Addi¬ tionally, while this description is written in terms of a thinly provisioned data container over and underlying file system, it should be noted that other thin provisioning imple¬ mentations may be utilized. As such, the use of an underlying file system to support a thinly provisioned data container should be taken as exemplary only. Accordingly this description is to be taken only by way of example and not to otherwise limit the scope of the invention. It is thus the object of the appended claims to cover all such variations and modifications as come within the true spirit and scope of the invention.

What is claimed is:

Claims

1 1. A method for reclaiming unused space from a thinly provisioned data container

2 served by a storage system, the method comprising the steps of:

3 determining one or more sets of blocks no longer in use on the thinly provisioned

4 data container;

5 sending a punch hole command to the storage system, the punch hole command β identifying one or more sets of blocks no longer in use on the thinly provisioned data

7 container; and

8 in response to receiving the punch hole command, deallocating the one or more

9 sets of blocks identified in the punch hole command.

i 2. The method of claim 1 wherein the data container comprises a virtual disk.

1 3. The method of claim 1 wherein the punch hole command comprises a small com-

2 puter systems interface vendor specific command.

1 4. The method of claim 1 wherein the punch hole command comprises an operation

2 code field, a number of ranges field and one or more range identifier fields.

1 5. The method of claim 4 wherein the range identifier fields comprise a logical block

2 address field and a range length field.

1 _a 6. The method of claim 1 wherein the punch hole command comprises an operation

2 code field, a bitmap size field, a logical block address field and a bitmap field.

1 7. The method of claim 1 wherein the step of determining one or more sets of blocks

2 no longer in use on the data container further comprises the step of querying a file system

3 on a client. 1 8. The method of claim 1 wherein the step of determining one or more sets of blocks

2 no longer in use on the data container further comprises the steps of:

₃ allocating a file on a file system overlaid onto the data container;

4 locking the allocated file;

5 identifying blocks storing the allocated file on the file system; and β identifying as the one or more sets of blocks no longer in use on the data container

7 the identified blocks storing the allocated file.

1 9. The method of claim 1 wherein the step of determining one or more sets of blocks

2 no longer in use on the data container comprises the step of querying a non-file system

3 application utilizing the data container.

1 10. The method of claim 9 wherein the non-file system application comprises a data-

2 base system.

1 11. A system for reclaiming unused space from a thinly provisioned data container

2 served by a storage system, the system comprising:

3 means for determining one or more sets of blocks no longer in use on the thinly

4 provisioned data container;

5 means for sending a punch hole command to the storage system, the punch hole

6 command identifying one or more sets of blocks no longer in use on the thinly provi-

7 sioned data container; and

8 means for in response to receiving the punch hole command, deallocating the one

9 or more sets of blocks identified in the punch hole command.

i 12. The system of claim 11 wherein the data container comprises a virtual disk.

1 13. The system of claim 11 wherein the punch hole command comprises a small com-

2 puter systems interface vendor specific command. 14. The system of claim 11 wherein the punch hole command comprises an operation code field, a number of ranges field and one or more range identifier fields.

15. The system of claim 14 wherein the range identifier fields comprise a logical block address field and a range length field.

16. The system of claim 11 wherein the punch hole command comprises an operation code field, a bitmap size field, a logical block address field and a bitmap field.

17. The system of claim 11 wherein the means for determining one or more sets of blocks no longer in use on the data container further comprises means for querying a file system on a client.

18. The system of claim 11 wherein the means for determining one or more sets of blocks no longer in use on the data container further comprises: means for allocating a file on a file system overlaid onto the data container; means for locking the allocated file; means for identifying blocks storing the allocated file on the file system; and means for identifying as the one or more sets of blocks no longer in use on the data container the identified blocks storing the allocated file.

19. The system of claim 11 wherein the means for determining one or more sets of blocks no longer in use on the data container further comprises means for querying a non- file system application utilizing the data container.

20. The system of claim 19 wherein the non-file system application comprises a data- base system.

21. A system for reclaiming unused space from a thinly provisioned data container served by a storage system, the system comprising: a host-side agent executing on a client of the storage system hosting the thinly provisioned data container, the host-side agent adapted to determine blocks of the data container that are no longer in use by the client and to send a punch hole command to the storage system, wherein the punch hole command identifies the blocks of the data con- tainer that are no longer in use by the client.

22. The system of claim 21 wherein the host-side agent is further adapted to query a file system on the client to determine blocks of the data container that are no longer in use by the client.

23. The system of claim 21 wherein the host-side agent is further adapted to query a non-file system application to determine blocks of the data container that are no longer in use by the client.

24. A computer readable medium for reclaiming unused space from a thinly provi- sioned data container served by a storage system, the computer readable medium includ- ing program instructions for performing the steps of: determining one or more sets of blocks no longer in use on the data container; sending a punch hole command to the storage system, the punch hole command identifying one or more sets of blocks no longer in use on the data container; and in response to receiving the punch hole command, de-allocating the one or more sets of blocks identified in the punch hole command.

25. A system for reclaiming unused space from a thinly provisioned data container served by a storage system, the system comprising: a storage operating system executing on the storage system, the storage operating system adapted to receive a punch hole command from a host-side agent and further adapted to deallocate one or more ranges of blocks identified in the punch hole com- mand.