262 results about "Live migration" patented technology

A transparent high-availability solution utilizing virtualization technology is presented. A cluster environment and management thereof is implemented through an automated installation and setup procedure resulting in a cluster acting as a single system. The cluster is setup in an isolated virtual machine on each of a number of physical nodes of the system. Customer applications are run within separate application virtual machines on one physical node at a time and are run independently and unaware of their configuration as part of a high-availability cluster. Upon detection of a failure, traffic is rerouted through a redundant node and the application virtual machines are migrated from the failing node to another node using live migration techniques.

Methods and apparatus are provided for controlling live migration of a virtual machine from a first host to a second host in a data center. A virtual machine manager may distribute to at least one host in a virtual network an updated mapping policy that maps a customer address of the virtual machine to a provider address of the migrated virtual machine. The updated mapping policy enables hosts in the virtual network to communicate with the migrated virtual machine. The updated mapping policy can be a shadow policy. The shadow policy is transmitted to hosts in the virtual network by the virtual machine manager before live migration of the virtual machine completes and is maintained by recipient hosts in an inactive state until triggered. The virtual machine manager notifies hosts in the virtual network to activate the shadow policy when live migration completes.

A method is provided to migrate a virtual machine from a source computing machine to a destination computing machine comprising: suspending transmission of requests from a request queue disposed in source computing machine memory associated with the VM from the request queue to a VF; while suspending the transmission of requests, determining when no more outstanding responses to prior requests remain to be received; in response to a determination that no more outstanding responses to prior requests remain to be received, transferring state information that is indicative of locations of requests inserted to the request queue from the VF to a PF and from the PF to a memory region associated with a virtualization intermediary of the source computing machine. After transferring the state information to source computing machine memory associated with a virtualization intermediary, resuming transmission of requests from locations of the request queue indicated by the state information to the PF; and transmitting the requests from the PF to the physical storage.

Datacenter clusters often employ live virtual machine (VM) migration to efficiently utilize cluster-wide resources. Gang migration refers to the simultaneous live migration of multiple VMs from one set of physical machines to another in response to events such as load spikes and imminent failures. Gang migration generates a large volume of network traffic and can overload the core network links and switches in a data center. The present technology reduces the network overhead of gang migration using global deduplication (GMGD). GMGD identifies and eliminates the retransmission of duplicate memory pages among VMs running on multiple physical machines in the cluster. A prototype GMGD reduces the network traffic on core links by up to 51% and the total migration time of VMs by up to 39% when compared to the default migration technique in QEMU / KVM, with reduced adverse performance impact on network-bound applications.

A system and method for performing a VM migration which manages a cluster of machines in a pool for live migration to the same feature set or behavior. In certain embodiments, machines within the pool can be configured to emulate a certain feature set to enable a VM migration amongst the similar pools. The emulation can be by either masking reporting of a feature set or enabling / disabling a feature set. The handling of emulation registers within the hardware occurs at a firmware level rather than an operating system or hypervisor level.

A memory state equivalency analysis fabric which notionally overlays a given compute cloud. Equivalent sections of memory state are identified, and that equivalency information is conveyed throughout the fabric. Such a compute cloud-wide memory equivalency fabric is utilized as a powerful foundation for numerous memory state management and optimization activities, such as workload live migration and memory de-duplication across the entire cloud.

A virtual machine is migrated between a source cloud and a destination cloud. A temporary storage is mounted to a partition management firmware of the source cloud and to a partition management firmware of the destination cloud. A first storage location migration migrates the virtual machine from the source cloud to the at least one temporary storage. A second storage location migration migrates the virtual machine from the at least one temporary storage to final destination storage of the destination cloud. The temporary storage is then unmounted from the partition management firmware of the source cloud and from the partition management firmware of the destination cloud.

In an embodiment of the invention, a method for secure live migration of a virtual machine (VM) in a virtualized computing environment can include selecting a VM in a secure virtualized computing environment for live migration to a different virtualized computing environment and blocking data communications with the selected VM and other VMs in the secure virtualized computing environment. The selected VM can be live migrated to the different virtualized computing environment and the VM cna be restarted in the different virtualized computing environment. Notably, a secure communicative link can be established between the restarted VM and at least one other of the VMs in the secure virtualized computing environment. Finally, data communications between the restarted VM and the at least one other of the VMs can be enabled over the secure communicative link.

MemX provides a distributed system that virtualizes cluster-wide memory to support data-intensive and large memory workloads in virtual machines (VMs), and provides benefits in virtualized settings: (1) VM workloads that access large datasets can perform low-latency I / O over virtualized cluster-wide memory; (2) VMs can transparently execute very large memory applications that require more memory than physical DRAM present in the host machine; (3) reduces the effective memory usage of the cluster by de-duplicating pages that have identical content; (4) existing applications do not require any modifications to benefit from MemX such as the use of special APIs, libraries, recompilation, or relinking; and (5) supports live migration of large-footprint VMs by eliminating the need to migrate part of their memory footprint resident on other nodes. Detailed evaluations show that large dataset applications and multiple concurrent VMs achieve significant performance improvements using MemX compared against virtualized local and iSCSI disks.

The invention provides a method and system for dynamic scheduling of virtual resources in a cloud computing network, the system comprises a cloud control server, a resource monitoring server, a resource scheduling server, at least one cluster controller and at least one node controller, and the method comprises: the resource monitoring server collects running information and sends the running information to the cloud control server; the node controller in which a virtual machine is positioned sends a real-time migration request to the cloud control server when achieving virtual resource bottleneck; the cloud control server sends the running information to the resource scheduling server; and the resource scheduling server searches the node controller with sufficient scheduling resources in the running information for performing real-time migration. The dynamic scheduling of the virtual resources is realized by adopting the real-time migration method, load balancing is dynamically realized, and the virtual resources in a cloud are utilized with high efficiency through the high-efficient load balancing.

Gang migration refers to the simultaneous live migration of multiple Virtual Machines (VMs) from one set of physical machines to another in response to events such as load spikes and imminent failures. Gang migration generates a large volume of network traffic and can overload the core network links and switches in a datacenter. In this paper, we present an approach to reduce the network overhead of gang migration using global deduplication (GMGD). GMGD identifies and eliminates the retransmission of duplicate memory pages among VMs running on multiple physical machines in the cluster. The design, implementation and evaluation of a GMGD prototype is described using QEMU / KVM VMs. Evaluations on a 30-node Gigabit Ethernet cluster having 10 GigE core links shows that GMGD can reduce the network traffic on core links by up to 65% and the total migration time of VMs by up to 42% when compared to the default migration technique in QEMU / KVM. Furthermore, GMGD has a smaller adverse performance impact on network-bound applications.

A first least recently used map is generated for a set of memory pages of a first virtual machine. The first least recently used map includes metadata including memory page physical address location information. A first of the memory pages of the first virtual machine and the metadata for the first memory page is sent from the first virtual machine to a second virtual machine while the first virtual machine is executing. A first memory page and meta data associated therewith of the second virtual machine is received from the second virtual machine at the first virtual machine. The memory pages of the first virtual machine are ordered from a first location of the first least recently used map to a last location of the first least recently used map based on how recently each of the memory pages of the first virtual machine has been used.

Techniques are disclosed for the non-disruptive and reliable live migration of a virtual machine (VM) from a source host to a target host, where network data is placed directly into the VM's memory. When a live migration begins, a network interface card (NIC) of the source stops placing newly received packets into the VM's memory. A virtual server driver (VSP) on the source stores the packets being processed and forces a return of the memory where the packets are stored to the NIC. When the VM has been migrated to the target, and the source VSP has transferred the stored packets to the target host, the VM resumes processing the packets, and when the VM sends messages to the target NIC that the memory associated with a processed packet is free, a VSP on the target intercepts that message, blocking the target NIC from receiving it.

A transparent high-availability solution utilizing virtualization technology is presented. A cluster environment and management thereof is implemented through an automated installation and setup procedure resulting in a cluster acting as a single system. The cluster is setup in an isolated virtual machine on each of a number of physical nodes of the system. Customer applications are run within separate application virtual machines on one physical node at a time and are run independently and unaware of their configuration as part of a high-availability cluster. Upon detection of a failure, traffic is rerouted through a redundant node and the application virtual machines are migrated from the failing node to another node using live migration techniques.

A method is provided to migrate a virtual machine from a source computing machine to a destination computing machine comprising: suspending transmission of requests from a request queue disposed in source computing machine memory associated with the VM from the request queue to a VF; while suspending the transmission of requests, determining when no more outstanding responses to prior requests remain to be received; in response to a determination that no more outstanding responses to prior requests remain to be received, transferring state information that is indicative of locations of requests inserted to the request queue from the VF to a PF and from the PF to a memory region associated with a virtualization intermediary of the source computing machine. After transferring the state information to source computing machine memory associated with a virtualization intermediary, resuming transmission of requests from locations of the request queue indicated by the state information to the PF; and transmitting the requests from the PF to the physical storage.

A method and apparatus for distribution of applications amongst a number of servers, ensuring that changes to application data on a master for that application are asynchronously replicated to a number of slaves for that application. Servers may be located in geographically diverse locations; the invention permits data replication over high-latency and lossy network connections and failure-tolerance under hardware and network failure conditions. Access to applications is mediated by a distributed protocol handler which allows any request for any application to be addressed to any server, and which, when working in tandem with the replication system, pauses connections momentarily to allow seamless, consistent live-migration of applications and their state between servers. Additionally, a system which controls the aforementioned live-migration based on dynamic measurement of load generated by each application and the topological preferences of each application, in order to automatically keep servers at an optimum utilisation level.

A computer implemented method for cloud bursting applied to a database includes employing a database live migration using a database hot backup capability to support database cloud bursting without client service interruption in multitenant databases, ascertaining a critical time to start / redo the database cloud bursting to a public cloud so as to meet a service level agreement SLA while leveraging a private cloud for the database to a maximum level, and determining tenants of the database to burst into the public cloud to responsive to costs and benefits.

The invention discloses a virtual machine live migration method, a virtual machine migration management apparatus, and a virtual machine live migration system. The method comprises following steps: the virtual machine migration management apparatus maps first mirror image files of a first virtual machine working on a source terminal physical machine to an object terminal physical machine, so that second mirror image files having correspondence relations with the first mirror image files can be generated on the object terminal physical machine; the virtual machine migration management apparatus sends migration commands to the source end physical machine; the virtual machine migration management apparatus controls a second virtual machine to work on the object terminal physical machine through calling memory data of the first virtual machine and the correspondence relations, wherein the memory data of the first virtual machine is stored in the object terminal physical machine, and the second virtual machine is pre-established on the object terminal physical machine and having the same configuration with the first virtual machine; and the virtual machine migration management apparatus synchronizes disk data in the first mirror image files to the second mirror image files when the second virtual machine works on the object terminal physical machine. According to the invention, a technical problem that the live migration time is long due to a pre-copy manner is solved.

A virtual machine system managed by a current host OS virtually operating on hardware is provided that activates a spare host OS by copying the current host OS to a prescribed memory device using a live migration function when the current host OS is activated, notifies the spare host OS of a request issued to the current host OS via a virtual machine monitor, changes a state of the spare host OS, and switches an OS for managing the virtual machine system from the current host OS to the spare host OS, when the current host OS is in an erroneous state.

The invention discloses a migrating method and device for a virtual machine, and partly solves the problem that the switch timing between a precopy and a post-copy in a live migration hybrid-copy method is hard to determine. In some of the applicable embodiment of the invention, the virtual machine migration method comprises the steps as follows: firstly, copying the memory data of a target virtual machine to be migrated on a source physical host to a target physical host in a circulatory iteration manner by adopting a precopy method; secondly, after each round of iteration copy is ended, calculating the changing rate of a dirty page, wherein the changing rate of the dirty page represents the rate of the number changing of the remaining internal memory dirty page in a current round of iteration copy relative to the remaining internal memory dirty page in a previous round of iteration copy; thirdly, judging whether the dirty page changing rate is within a dirty page changing rate threshold range or not; fourthly, counting the frequency that the dirty page changing rate is within the dirty page changing rate threshold range and judging whether the frequency is larger than or equal to the threshold value; finally, if the frequency is larger than or equal to the threshold value, adopting the post-copy method for migrating the target virtual machine to the target physical host.

A system, method, computer program, and / or computer readable medium for providing application isolation and live migration to one or more applications and their associated resources. The computer readable medium includes computer-executable instructions for execution by a processing system. The computer-executable instructions may be for creating a new isolated environment, for creating a copy of an isolated environment, live migrating applications within and between isolated environments, and for deploying an isolated environment Further, the instructions may be for launching one or more applications within one or more isolated environment, for terminating one or more applications within one or more isolated environments, for configuring resource mappings, exceptions and policies, for triggering a live migrate, and for delivery of one or more applications within isolated environments over the network.

The invention provides a distributed file system data volume deployment method based on a virtual address pool. A node cluster or a sub-cluster is built according to different node roles of metadata nodes, data storage nodes and client side nodes in a distributed file system; a data storage volume is deployed in a mode based on the virtual IP address pool on a data node cluster to separate the deployment of a distributed file system data storage volume from physical addresses of data nodes so as to combine the deployment of the data storage volume with virtual addresses of the data node cluster; a monitoring module is deployed on an operating node of the distributed file system so as to realize real-time monitoring of the operating conditions of the node and a load; according to the conditions of the node and the load, a virtual IP borne by any node and the data storage volume mounted on one virtual address are transferred in real time in the data node cluster; meanwhile, a failure of a physical node is also isolated, so that the effect that data services of the data storage volume are not interrupted when the physical node is maintained is ensured.

A mechanism for controlling live migration of a running application between a source machine and a target machine. At least one application characteristic and at least one network characteristic are obtained. An objective function that defines a cost to migrate from the source machine to the target machine as a function such as a weighted combination of at least one cost property is obtained. The objective function is optimized using the at least one application characteristic and the at least one network characteristic to determine a policy specifying when to switch execution of the running application from the source machine to the target machine. Live migration of the application from the source machine is controlled to send pages from the source machine to the target machine in accordance with the policy, and to switch the execution of the application from the source machine to the target machine at a state specified by the policy.

A mechanism for guaranteeing deterministic bounded tunable downtime for live migration of virtual machines (VMs) over reliable channels is disclosed. A method of the invention includes receiving a maximum downtime of a VM during a live migration of the VM from a source host machine to a destination host machine, determining a capacity of a network channel on which the VM is live migrated to the destination host machine, calculating a number of memory pages of the VM left to be transferred as part of the live migration before the VM is stopped, wherein the calculating of the number of memory pages is based on the received maximum downtime and the determined capacity of the network channel, and stopping the VM when the calculated number of memory pages remain to be transferred in order to transfer those memory pages to the destination host machine.

Methods and systems for database migration from a multitenant database include taking a snapshot of an original database to be migrated with a hot backup process, such that the database is still capable of answering queries during the hot backup process; maintaining a query log of all queries to the tenant database after the hot backup process begins; initializing a new database at a target server using the snapshot; replaying the query log synchronize the new database with the original database; and answering new queries with the new database and not the original database.

Methods and systems for database migration from a multitenant database include taking a snapshot of an original database to be migrated with a hot backup process, such that the database is still capable of answering queries during the hot backup process; maintaining a query log of all queries to the tenant database after the hot backup process begins; initializing a new database at a target server using the snapshot; replaying the query log synchronize the new database with the original database; and answering new queries with the new database and not the original database.

A partition migration mechanism migrates a logical partition executing an operating system and resumes the logical partition before all resources in the logical partition have been migrated. When a partition is being migrated, a call checkpoint mechanism creates checkpoints of the state of the operating system when the partition manager is called. Before performing the call to the partition manager, a check is made to determine if all resources required by the call are available. If so, the partition manager call is executed. If all resources required by the call are not available, a resource fault is indicated, which causes the operating system state from the last checkpoint to be restored and a corresponding virtual CPU to be preempted until the resource that caused the fault becomes available. Exceptions that do not require the missing resource may be performed while the virtual CPU awaits the resource to become available.

A computer-implemented method is provided for reducing network traffic associated with live migration. The method comprises selecting a virtual machine on a first physical host for migration to a second physical host, and providing each application running in the selected virtual machine with notification that the migration has started. In response to receiving the notification that the migration has started, each application reducing the number of memory pages in use during the migration. The selected virtual machine and each application is then migrated from the first physical host to the second physical host, wherein the migration of each application includes migration of the reduced number of memory pages. Non-limiting examples of methods that may be used to reduce the number of memory pages includes identifying memory that contains garbage, reducing the size of the heap, increasing the pause time used for garbage collection, deferring just-in-time compilation of hot methods, and releasing caches and buffers.

A virtual machine control method and a virtual machine having the dual objectives of utilizing NIC on a virtual machine that creates sub-virtual machines operated by a VMM on virtual machines generated by a hypervisor to avoid software copying by the VMM and to prevent band deterioration during live migration or adding sub-virtual machines. In a virtual machine operating plural virtualization software on a physical machine including a CPU, memory, and multi-queue NIC; a virtual multi-queue NIC is loaded in the virtual machine, for virtual queues included in the virtual multi-queue NIC, physical queues configuring the multi-queue NIC are assigned to virtual queues where usage has started, and the physical queues are allowed direct access to the virtual machine memory.