A “virtual machine” or a “VM” refers to a specific software-based implementation of a machine in a virtualization environment, in which the hardware resources of a real computer (e.g., CPU, memory, etc.) are virtualized or transformed into the underlying support for the fully functional virtual machine that can run its own operating system and applications on the underlying physical resources just like a real computer.
Virtualization works by inserting a thin layer of software directly on the computer hardware or on a host operating system. This layer of software contains a virtual machine monitor or “hypervisor” that allocates hardware resources dynamically and transparently. Multiple operating systems run concurrently on a single physical computer and share hardware resources with each other. By encapsulating an entire machine, including CPU, memory, operating system, and network devices, a virtual machine is completely compatible with most standard operating systems, applications, and device drivers. Most modern implementations allow several operating systems and applications to safely run at the same time on a single computer, with each having access to the resources it needs when it needs them.
Virtualization allows multiple virtual machines to run on a single physical machine, with each virtual machine sharing the resources of that one physical computer across multiple environments. Different virtual machines can run different operating systems and multiple applications on the same physical computer.
One reason for the broad adoption of virtualization in modern business and computing environments is because of the resource utilization advantages provided by virtual machines. Without virtualization, if a physical machine is limited to a single dedicated operating system, then during periods of inactivity by the dedicated operating system the physical machine is not utilized to perform useful work. This is wasteful and inefficient if there are users on other physical machines which are currently waiting for computing resources. To address this problem, virtualization allows multiple VMs to share the underlying physical resources so that during periods of inactivity by one VM, other VMs can take advantage of the resource availability to process workloads. This can produce great efficiencies for the utilization of physical devices, and can result in reduced redundancies and better resource cost management.
Many organizations use data centers to implement virtualization, where the data centers are typically architected with traditional servers that communicate with a set of networked storage devices over a network. For example, many data centers are designed using diskless computers (“application servers”) that communicate with a set of networked storage appliances (“storage servers”) via a network, such as a Fiber Channel or Ethernet network.
The problem is that this traditional approach cannot adapt to the modern demands of virtualization, which is particularly problematic with respect to the way these traditional architectures manage storage. One reason for this is because the traditional network storage-based architecture is designed for physical servers that serve relatively static workloads, but which is not flexible or adaptable enough to adequately handle the dynamic nature of storage and virtual machines that, in a virtualization or cloud computing environment, may be created or moved on the fly from one network location to another.
Moreover, the traditional approach relies upon very large and specialized rackmount or freestanding compute and storage devices that are managed by a central storage manager. This approach does not scale very well, since the central storage manager becomes a very significant performance bottleneck as the number of storage devices increase. Moreover, the traditional compute and storage devices are expensive to purchase, maintain, and power, and are large enough to require a significant investment just in terms of the amount of physical space that is needed to implement the data center.
Given these challenges with the traditional data center architectures, it has become clear that the conventional approaches to implement a data center for virtualization presents excessive levels of cost and complexity, while being very ill-adapted to the needs of modern virtualization systems. These problems are further exacerbated by the fact that data volumes are constantly growing at a rapid pace in the modern data center, thanks to the ease of creating new VMs. In the enterprise, new initiatives like desktop virtualization contribute to this trend of increased data volumes. This growing pool of VMs is exerting tremendous cost, performance and manageability pressure on the traditional architecture that connects compute to storage over a multi-hop network.
Therefore, there is a need for an improved approach to implement an architecture for a virtualization data center.
Embodiments of the present invention provide an improved architecture which enables significant convergence of the components of a system to implement virtualization. The infrastructure is VM-aware, and permits SOCS provisioning to allow storage on a per-VM basis, while identifying I/O coming from each VM. The current approach can scale out from a few nodes to a large number of nodes. In addition, the inventive approach has ground-up integration with all types of storage, including solid-state drives. The architecture of the invention provides high availability against any type of failure, including disk or node failures. In addition, the invention provides high performance by making I/O access local, leveraging solid-state drives and employing a series of patent-pending performance optimizations.
Further details of aspects, objects, and advantages of the invention are described below in the detailed description, drawings, and claims. Both the foregoing general description and the following detailed description are exemplary and explanatory, and are not intended to be limiting as to the scope of the invention.
The drawings illustrate the design and utility of embodiments of the present invention, in which similar elements are referred to by common reference numerals. In order to better appreciate the advantages and objects of embodiments of the invention, reference should be made to the accompanying drawings. However, the drawings depict only certain embodiments of the invention, and should not be taken as limiting the scope of the invention.
Embodiments of the present invention provide an improved approach to implement virtualization appliances for a datacenter which address and correct the problems of the prior art. According to some embodiments, the present invention provides a scalable compute and storage infrastructure that effectively and efficiently allows organizations to virtualize their data centers. The virtualization appliance of the present invention provides complete compute and storage capabilities along with performance, scalability, availability and data management features. In some embodiments, the virtualization appliances leverages industry-standard hardware components and advanced storage management software to provide an out-of-the-box solution that makes virtualization extremely easy and cost effective.
In addition, local storage from all nodes is virtualized into a unified storage pool, which is referred to herein as “scale-out converged storage” or “SOCS” 155. As described in more detail below, SOCS 155 acts like an advanced SAN that uses local SSDs and disks from all nodes to store virtual machine data. Virtual machines running on the cluster write data to SOCS as if they were writing to a SAN. SOCS is VM-aware and provides advanced data management features. This approach brings the data closer to virtual machines by storing the data locally on the system (if desired), resulting in higher performance at a lower cost. As discussed in more detail below, this solution can horizontally scale from a few nodes to a large number of nodes, enabling organizations to scale their infrastructure as their needs grow.
While traditional SAN solutions typically have 1, 2, 4 or 8 controllers, an n-node system according to the present embodiment has n controllers. Every node in the cluster runs a special virtual machine, called a Controller VM (or “service VM”), which acts as a virtual controller for SOCS. All Controller VMs in the cluster communicate with each other to form a single distributed system. Unlike traditional SAN/NAS solutions that are limited to a small number of fixed controllers, this architecture continues to scale as more nodes are added.
As stated above, each block includes a sufficient collection of hardware and software to provide a self-contained virtualization appliance, e.g., as shown in
Each node in the block includes both hardware components 202 and software components 204 to implement virtualization. Hardware components 202 includes processing capacity (e.g., using one or more processors) and memory capacity (e.g., random access memory or RAM) on a motherboard 203. The node also comprises local storage 222, which in some embodiments include Solid State Drives (henceforth “SSDs”) 125 and/or Hard Disk Drives (henceforth “HDDs” or “spindle drives”) 127. Any combination of SSDs and HDDs may be used to implement the local storage 222.
The software 204 includes a hypervisor 230 to manage the interactions between the underlying hardware 202 and the one or more user VMs 202a and 202b that run client software. A controller VM 210a exists on each node to implement distributed storage management of the local storage 222, such that the collected local storage for all nodes can be managed as a combined SOCS.
Virtual disks (or “vDisks”) can be structured from the storage devices in the storage pool 360, as described in more detail below. As used herein, the term vDisk refers to the storage abstraction that is exposed by a Controller VM to be used by a user VM. In some embodiments, the vDisk is exposed via iSCSI (“internet small computer system interface”) or NFS (“network file system”) and is mounted as a virtual disk on the user VM. Each server 300a or 300b runs virtualization software, such as VMware ESX(i), Microsoft Hyper-V, or RedHat KVM. The virtualization software includes a hypervisor 330/332 to manage the interactions between the underlying hardware and the one or more user VMs 302a, 302b, 302c, and 302d that run client software.
Controller VM 310a/310b (also referred to herein as “service VMs”) are used to manage storage and I/O activities. This is the distributed “Storage Controller” in the currently described architecture. Multiple such storage controllers coordinate within a cluster to form a single-system. The Controller VMs 310a/310b are not formed as part of specific implementations of hypervisors 330/332. Instead, the Controller VMs run as virtual machines above hypervisors 330/332 on the various nodes/servers 302a and 302b, and work together to form a distributed system 310 that manages all the storage resources, including the locally attached storage 322/324, the networked storage 328, and the cloud storage 326. Since the Controller VMs run above the hypervisors 330/332, this means that the current approach can be used and implemented within any virtual machine architecture, since the Controller VMs of embodiments of the invention can be used in conjunction with any hypervisor from any virtualization vendor.
Each Controller VM 310a-b exports one or more block devices or NFS server targets that appear as disks to the client VMs 302a-d. These disks are virtual, since they are implemented by the software running inside the Controller VMs 310a-b. Thus, to the user VMs 302a-d, the Controller VMs 310a-b appear to be exporting a clustered storage appliance that contains some disks. All user data (including the operating system) in the client VMs 302a-d resides on these virtual disks.
Significant performance advantages can be gained by allowing the virtualization system to access and utilize local (e.g., server-internal) storage 322 as disclosed herein. This is because I/O performance is typically much faster when performing access to local storage 322 as compared to performing access to networked storage 328 across a network 340. This faster performance for locally attached storage 322 can be increased even further by using certain types of optimized local storage devices, such as SSDs 325. Once the virtualization system is capable of managing and accessing locally attached storage, as is the case with the present embodiment, various optimizations can then be implemented to improve system performance even further. For example, the data to be stored in the various storage devices can be analyzed and categorized to determine which specific device should optimally be used to store the items of data. Data that needs to be accessed much faster or more frequently can be identified for storage in the locally attached storage 322. On the other hand, data that does not require fast access or which is accessed infrequently can be stored in the networked storage devices 328 or in cloud storage 326. In addition, the performance of the local storage can be further improved by changing the mix of SSDs and HDDs within the local storage, e.g., by increasing or decreasing the proportion of SSDs to HDDs in the local storage.
The present architecture solves storage challenges for virtual machines providing a general-purpose scale-out compute and storage infrastructure that eliminates the need for network storage. In part, this is due to the distributed nature of the storage controller infrastructure that utilizes controller VMs to act as a virtual controller for SOCS. Since all the Controller VMs in the cluster communicate with each other to form a single distributed system, this eliminates the limitations and performance bottlenecks associated with traditional SAN solutions that typically have only 1, 2, 4 or 8 controllers. Therefore, n-node clusters will essentially have n controllers, providing a solution that will easily scale to very large data volumes.
In addition, the solution will very effectively support virtualization and hypervisor functions, within a single virtualization appliance (block) that can be extensively combined with other blocks to support large scale virtualization needs. Since the architecture is VM-aware, it overcomes limitations of traditional solutions that were optimized to work only with physical servers. For example, the present approach overcomes limitations associated with the traditional unit of management for storage pertaining to LUNs, where if a LUN is shared by many VMs, it becomes more difficult to perform storage operations such as backup, recovery, and snapshots on a per-VM basis. It is also difficult to identify performance bottlenecks in a heavily-shared environment due to the chasm between computing and storage tiers. The current architecture overcomes these limitations since the storage units (vdisks) are managed across an entire virtual storage space.
Moreover, the present approach can effectively take advantage of enterprise-grade solid-state drives (SSDs). Traditional storage systems were designed for spinning media and it is therefore difficult for these traditional systems to leverage SSDs efficiently due to the entirely different access patterns that SSDs provide. While hard disks have to deal with the rotation and seek latencies, SSDs do not have such mechanical limitations. This difference between the two media requires the software to be optimized differently for performance. One cannot simply take software written for hard disk-based systems and hope to use it efficiently on solid-state drives. The present architecture can use any type of storage media, including SSDs, and can use SSDs to store a variety of frequently-accessed data, from VM metadata to primary data storage, both in a distributed cache for high-performance and in persistent storage for quick retrieval.
In some embodiment, to maximize the performance benefits of using SSDs, the present architecture reserves SSDs for 1/0-intensive functions and includes space-saving techniques that allow large amounts of logical data to be stored in a small physical space. In addition, the present approach can be used to migrate “cold” or infrequently-used data to hard disk drives automatically, allowing administrators to bypass SSDs for low-priority VMs.
The present architecture therefore provides a solution that enables significant convergence of the storage components of the system with the compute components, allowing VMs and SOCS to co-exist within the same cluster. From a hardware perspective, each block provides a “building block” to implement an expandable unit of virtualization, which is both self-contained and expandable to provide a solution for any sized requirements.
Each of the serverboards 406 acts as a separate node within the block 400. As independent nodes, each node may be powered on or off separately without affecting the others. In addition, the serverboards 406 are hot swappable and may be removed from the end of the chassis without affecting the operation of the other serverboards. This configuration of multiple nodes ensures hardware-based redundancy of processing and storage capacity for the block, with the storage management software providing for operational redundancies of the data stored and managed by the block.
The block 400 also includes multiple power supply modules 408, e.g., two separate modules as shown in
The block 400 supports multiple local storage devices. In some embodiments, the block 400 includes a backplane that allows connection of six SAS or SATA storage units to each node, for a total of 24 storage units 404 for the block 400. Any suitable type or configuration of storage unit may be connected to the backplane, such as SSDs or HDDs. In some embodiments, any combination of SSDs and HDDs can be implemented to form the six storage units for each node, including all SSDs, all HDDs, or a mixture of SSDs and HDDs.
The entirety of the block 400 fits within a “2u” or less form factor unit. A rack unit or “u” (also referred to as a “RU”) is a unit of measure used to describe the height of equipment intended for mounting in a rack system. In some embodiments, one rack unit is 1.75 inches (44.45 mm) high. This means that the 2u or less block provides a very space-efficient and power-efficient building block for implementing a virtualized data center. The redundancies that are built into the block mean that there is no single point of failure that exists for the unit. The redundancies also mean that there is no single point of bottleneck for the performance of the unit.
The blocks are rackable as well, with the block being mountable on a standard 19″ rack.
Here, the user VM 702 structures its I/O requests into the iSCSI format. The iSCSI or NFS request 750a designates the IP address for a Controller VM from which the user VM 702 desires I/O services. The iSCSI or NFS request 750a is sent from the user VM 702 to a virtual switch 752 within hypervisor 752 to be routed to the correct destination. If the request is to be intended to be handled by the Controller VM 710a within the same server 700a, then the iSCSI or NFS request 750a is internally routed within server 700a to the Controller VM 710a. As described in more detail below, the Controller VM 710a includes structures to properly interpret and process that request 750a.
It is also possible that the iSCSI or NFS request 750a will be handled by a Controller VM 710b on another server 700b. In this situation, the iSCSI or NFS request 750a will be sent by the virtual switch 752 to a real physical switch to be sent across network 740 to the other server 700b. The virtual switch 755 within the hypervisor 733 on the server 733 will then route the request 750a to the Controller VM 710b for further processing.
According to some embodiments, the controller VM runs the Linux operating system. As noted above, since the controller VM exports a block-device or file-access interface to the user VMs, the interaction between the user VMs and the controller VMs follows the iSCSI or NFS protocol, either directly or indirectly via the hypervisor's hardware emulation layer.
For easy management of the appliance, the Controller VMs all have the same IP address isolated by internal VLANs (virtual LANs in the virtual switch of the hypervisor).
The second virtual NIC 761b is used to communicate with entities external to the node 700a, where the virtual NIC 761b is associated with an IP address that would be specific to Controller VM 710a (and no other controller VM). The second virtual MC 761b is therefore used to allow Controller VM 710a to communicate with other controller VMs, such as Controller VM 710b on node 700b. It is noted that Controller VM 710b would likewise utilize VLANs and multiple virtual NICs 763a and 763b to implement management of the appliance.
For easy management of the appliance, the storage is divided up into abstractions that have a hierarchical relationship to each other.
Storage with similar characteristics is classified into tiers. Thus, all SSDs can be classified into a first tier and all HDDs may be classified into another tier etc. In a heterogeneous system with different kinds of HDDs, one may classify the disks into multiple HDD tiers. This action may similarly be taken for SAN and cloud storage.
The storage universe is divided up into storage pools—essentially a collection of specific storage devices. An administrator may be responsible for deciding how to divide up the storage universe into storage pools. For example, an administrator may decide to just make one storage pool with all the disks in the storage universe in that pool. However, the principal idea behind dividing up the storage universe is to provide mutual exclusion—fault isolation, performance isolation, administrative autonomy—when accessing the disk resources.
This may be one approach that can be taken to implement QoS techniques. For example, one rogue user may result in an excessive number of random IO activity on a hard disk—thus if other users are doing sequential IO, they still might get hurt by the rogue user. Enforcing exclusion (isolation) through storage pools might be used to provide hard guarantees for premium users. Another reason to use a storage pool might be to reserve some disks for later use (field replaceable units, or “FRUs”).
As noted above, the Controller VM is the primary software component within the server that virtualizes I/O access to hardware resources within a storage pool according to embodiments of the invention. This approach essentially provides for a separate and dedicated controller for each and every node within a virtualized data center (a cluster of nodes that run some flavor of hypervisor virtualization software), since each node will include its own Controller VM. This is in contrast to conventional storage architectures that provide for a limited number of storage controllers (e.g., four controllers) to handle the storage workload for the entire system, and hence results in significant performance bottlenecks due to the limited number of controllers. Unlike the conventional approaches, each new node will include a Controller VM to share in the overall workload of the system to handle storage tasks. Therefore, the current approach is infinitely scalable, and provides a significant advantage over the conventional approaches that have a limited storage processing power. Consequently, the currently described approach creates a massively-parallel storage architecture that scales as and when hypervisor hosts are added to a datacenter.
The main entry point into the Controller VM is the central controller module 804 (which is referred to here as the “I/O Director module 804”). The term I/O Director module is used to connote that fact that this component directs the I/O from the world of virtual disks to the pool of physical storage resources. In some embodiments, the I/O Director module implements the iSCSI or NFS protocol server.
A write request originating at a user VM would be sent to the iSCSI or NFS target inside the controller VM's kernel. This write would be intercepted by the I/O Director module 804 running in user space. I/O Director module 804 interprets the iSCSI LUN or the NFS file destination and converts the request into an internal “vDisk” request (e.g., as described in more detail below). Ultimately, the I/O Director module 804 would write the data to the physical storage.
Each vDisk managed by a Controller VM corresponds to a virtual address space forming the individual bytes exposed as a disk to user VMs. Thus, if the vDisk is of size 1 TB, the corresponding address space maintained by the invention is 1 TB. This address space is broken up into equal sized units called vDisk blocks. Metadata 810 is maintained by the Controller VM to track and handle the vDisks and the data and storage objects in the system that pertain to the vDisks. The Metadata 810 is used to track and maintain the contents of the vDisks and vDisk blocks.
In order to determine where to write and read data from the storage pool, the I/O Director module 804 communicates with a Distributed Metadata Service module 830 that maintains all the metadata 810. In some embodiments, the Distributed Metadata Service module 830 is a highly available, fault-tolerant distributed service that runs on all the Controller VMs in the appliance. The metadata managed by Distributed Metadata Service module 830 is itself kept on the persistent storage attached to the appliance. According to some embodiments of the invention, the Distributed Metadata Service module 830 may be implemented on SSD storage.
Since requests to the Distributed Metadata Service module 830 may be random in nature, SSDs can be used on each server node to maintain the metadata for the Distributed Metadata Service module 830. The Distributed Metadata Service module 830 stores the metadata that helps locate the actual content of each vDisk block. If no information is found in Distributed Metadata Service module 830 corresponding to a vDisk block, then that vDisk block is assumed to be filled with zeros. The data in each vDisk block is physically stored on disk in contiguous units called extents. Extents may vary in size when de-duplication is being used. Otherwise, an extent size coincides with a vDisk block. Several extents are grouped together into a unit called an extent group. An extent group is then stored as a file on disk. The size of each extent group is anywhere from 16 MB to 64 MB. In some embodiments, an extent group is the unit of recovery, replication, and many other storage functions within the system.
Further details regarding methods and mechanisms for implementing a Controller VM are described below and in U.S. Pat. No. 8,601,473, issued on Dec. 3, 2013, which is hereby incorporated by reference in its entirety. Further details regarding methods and mechanisms for implementing Metadata 910 are described below and in U.S. Pat. No. 8,850,130, issued on Sep. 30, 2014, which is hereby incorporated by reference in its entirety.
A health management module 808 (which may hereinafter be referred to as a “Curator”) is employed to address and cure any inconsistencies that may occur with the Metadata 810. The Curator 808 oversees the overall state of the virtual storage system, and takes actions as necessary to manage the health and efficient performance of that system. According to some embodiments of the invention, the curator 808 operates on a distributed basis to manage and perform these functions, where a master curator on a first server node manages the workload that is performed by multiple slave curators on other server nodes. MapReduce operations are performed to implement the curator workload, where the master curator may periodically coordinate scans of the metadata in the system to manage the health of the distributed storage system. Further details regarding methods and mechanisms for implementing Curator 808 are disclosed in U.S. Pat. No. 8,549,518, issued on Oct. 1, 2013, which is hereby incorporated by reference in its entirety.
Some of the Controller VMs also includes a Distributed Configuration Database module 806 to handle certain administrative tasks. The primary tasks performed by the Distributed Configuration Database module 806 are to maintain configuration data 812 for the Controller VM and act as a notification service for all events in the distributed system. Examples of configuration data 812 include, for example, (1) the identity and existence of vDisks; (2) the identity of Controller VMs in the system; (3) the physical nodes in the system; and (4) the physical storage devices in the system. For example, assume that there is a desire to add a new physical disk to the storage pool. The Distributed Configuration Database module 806 would be informed of the new physical disk, after which the configuration data 812 is updated to reflect this information so that all other entities in the system can then be made aware for the new physical disk. In a similar way, the addition/deletion of vDisks, VMs and nodes would handled by the Distributed Configuration Database module 806 to update the configuration data 812 so that other entities in the system can be made aware of these configuration changes.
Another task that is handled by the Distributed Configuration Database module 806 is to maintain health information for entities in the system, such as the Controller VMs. If a Controller VM fails or otherwise becomes unavailable, then this module tracks this health information so that any management tasks required of that failed Controller VM can be migrated to another Controller VM.
The Distributed Configuration Database module 806 also handles elections and consensus management within the system. Another task handled by the Distributed Configuration Database module is to implement ID creation. Unique IDs are generated by the Distributed Configuration Database module as needed for any required objects in the system, e.g., for vDisks, Controller VMs, extent groups, etc. In some embodiments, the IDs generated are 64-bit IDs, although any suitable type of IDs can be generated as appropriate for embodiment so the invention. According to some embodiments of the invention, the Distributed Configuration Database module 806 may be implemented on an SSD storage because of the real-time guarantees required to monitor health events.
The vDisks can either be unshared (read and written by a single user VM) or shared (accessed by multiple user VMs or hypervisors) according to embodiments of the invention.
For I/O requests 950b from a user VM 902b that resides on the same server node 900b, the process to handle the I/O requests 950b is straightforward, and is conducted as described above. Essentially, the I/O request is in the form of an iSCSI or NFS request that is directed to a given IP address. The IP address for the I/O request is common for all the Controller VM on the different server nodes, but VLANs allows the IP address of the iSCSI or NFS request to be private to a particular (local) subnet, and hence the I/O request 950b will be sent to the local Controller VM 910b to handle the I/O request 950b. Since local Controller VM 910b recognizes that it is the owner of the vDisk 923 which is the subject of the I/O request 950b, the local Controller VM 910b will directly handle the I/O request 950b.
Consider the situation if a user VM 902a on a server node 900a issues an I/O request 950a for the shared vDisk 923, where the shared vDisk 923 is owned by a Controller VM 910b on a different server node 900b. Here, the I/O request 950a is sent as described above from the user VM 902a to its local Controller VM 910a. However, the Controller VM 910a will recognize that it is not the owner of the shared vDisk 923. Instead, the Controller VM 910a will recognize that Controller VM 910b is the owner of the shared vDisk 923. In this situation, the I/O request will be forwarded from Controller VM 910a to Controller VM 910b so that the owner (Controller VM 910b) can handle the forwarded I/O request. To the extent a reply is needed, the reply would be sent to the Controller VM 910a to be forwarded to the user VM 902a that had originated the I/O request 950a.
In some embodiments, an IP table 902 (e.g., a network address table or “NAT”) is maintained inside the Controller VM 910a. The IP table 902 is maintained to include the address of the remote Server VMs. When the local Controller VM 910a recognizes that the I/O request needs to be sent to another Controller VM 910b, the IP table 902 is used to look up the address of the destination Controller VM 910b. This “NATing” action is performed at the network layers of the OS stack at the Controller VM 910a, when the local Controller VM 910a decides to forward the IP packet to the destination Controller VM 910b.
Each un-shared vDisk is owned by the Controller VM that is local to the user VM which accesses that vDisk on the shared-nothing basis. In the current example, vDisk 1023a is owned by Controller VM 1010a since this Controller VM is on the same server node 1000a as the user VM 1002a that accesses this vDisk. Similarly, vDisk 1023b is owned by Controller VM 1010b since this Controller VM is on the same server node 1000b as the user VM 1002b that accesses this vDisk.
I/O requests 1050a that originate user VM 1002a would therefore be handled by its local Controller VM 1023a on the same server node 1000a. Similarly, I/O requests 1050b that originate user VM 1002b would therefore be handled by its local Controller VM 1023b on the same server node 1000b. This is implemented using the same approach previously described above, in which the I/O request in the form of an iSCSI or NFS request is directed to a given IP address, and where VLANs allows the IP address of the iSCSI or NFS request to be private to a particular (local) subnet where the I/O request 950b will be sent to the local Controller VM to handle the I/O request. Since local Controller VM recognizes that it is the owner of the vDisk which is the subject of the I/O request, the local Controller VM will directly handle the I/O request.
It is possible that a user VM will move or migrate from one node to another node. Various virtualization vendors have implemented virtualization software that allows for such movement by user VMs. For shared vDisks, this situation does not necessarily affect the configuration of the storage system, since the I/O requests will be routed to the owner Controller VM of the shared vDisk regardless of the location of the user VM. However, for unshared vDisks, movement of the user VMs could present a problem since the I/O requests are handled by the local Controller VMs.
Therefore, what has been described is an improved architecture that enables significant convergence of the components of a system to implement virtualization. The infrastructure is VM-aware, and permits SOCS provisioning to allow storage on a per-VM basis, while identifying I/O coming from each VM. The current approach can scale out from a few nodes to a large number of nodes. In addition, the inventive approach has ground-up integration with all types of storage, including solid-state drives. The architecture of the invention provides high availability against any type of failure, including disk or node failures. In addition, the invention provides high performance by making I/O access local, leveraging solid-state drives and employing a series of patent-pending performance optimizations.
System Architecture
According to one embodiment of the invention, computer system 1400 performs specific operations by processor 1407 executing one or more sequences of one or more instructions contained in system memory 1408. Such instructions may be read into system memory 1408 from another computer readable/usable medium, such as static storage device 1409 or disk drive 1410. In alternative embodiments, hard-wired circuitry may be used in place of or in combination with software instructions to implement the invention. Thus, embodiments of the invention are not limited to any specific combination of hardware circuitry and/or software. In one embodiment, the term “logic” shall mean any combination of software or hardware that is used to implement all or part of the invention.
The term “computer readable medium” or “computer usable medium” as used herein refers to any medium that participates in providing instructions to processor 1407 for execution. Such a medium may take many forms, including but not limited to, non-volatile media and volatile media. Non-volatile media includes, for example, optical or magnetic disks, such as disk drive 1410. Volatile media includes dynamic memory, such as system memory 1408.
Common forms of computer readable media includes, for example, floppy disk, flexible disk, hard disk, magnetic tape, any other magnetic medium, CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with patterns of holes, RAM, PROM, EPROM, FLASH-EPROM, any other memory chip or cartridge, or any other medium from which a computer can read.
In an embodiment of the invention, execution of the sequences of instructions to practice the invention is performed by a single computer system 1400. According to other embodiments of the invention, two or more computer systems 1400 coupled by communication link 1415 (e.g., LAN, PTSN, or wireless network) may perform the sequence of instructions required to practice the invention in coordination with one another.
Computer system 1400 may transmit and receive messages, data, and instructions, including program, i.e., application code, through communication link 1415 and communication interface 1414. Received program code may be executed by processor 1407 as it is received, and/or stored in disk drive 1410, or other non-volatile storage for later execution.
In the foregoing specification, the invention has been described with reference to specific embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention. For example, the above-described process flows are described with reference to a particular ordering of process actions. However, the ordering of many of the described process actions may be changed without affecting the scope or operation of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative rather than restrictive sense.
The present application is a continuation application of U.S. patent application Ser. No. 13/551,291, filed on Jul. 17, 2012, issued as U.S. Pat. No. 9,772,866 on Sep. 26, 2017, which is hereby incorporated by reference in its entirety.
Number | Date | Country | |
---|---|---|---|
Parent | 13551291 | Jul 2012 | US |
Child | 15714880 | US |