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 data center. 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. We describe the design, implementation and evaluation of a GMGD prototype using QEMU/KVM VMs. Evaluations on a 30-node Gigabit Ethernet cluster having 10GigE 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.

/pdf/gang-migration-of-virtual-machines-using-cluster-wide-55krh066a5.pdf

Gang migration of virtual machines using cluster-wide deduplication

The purpose of the present invention is to enable correct stream reception to take place at the receiving end, in a manner accurately adapted to dynamic changes in delivered content. In a first transmission mode, a predetermined number of extended video streams constituting scalable-coded image data, and respectively including a basic video stream that includes lowest-layer image data, and image data of layers other than the lowest layer. In a second transmission mode, only the basic video stream that includes basic image data is transmitted. Identifying information for identifying the first mode and the second mode is inserted into the video stream. At the receiving end, the transmission mode can be easily ascertained, making it possible for correct stream reception to take place in a manner accurately adapted to changes in the stream composition, that is, dynamic changes in delivered content.

Transmission device, transmission method, reception device, and reception method

A recipient node of a multi-node data partitioning landscape can receive, directly from a requesting machine without being handled by a master node, a first data request related to a table. A target node of a plurality of processing nodes can be identified to handle the data request. The determining can include the recipient node applying partitioning information to determine a target data partition of the plurality of data partitions to which the data request should be directed and mapping information associating each data partition of the plurality of data partitions with an assigned node of the plurality of processing nodes. The recipient node can redirect the data request to the target node so that the target node can act on the target data partition in response to the data request.

Independent Table Nodes In Parallelized Database Environments

Application Delivery Controller (ADC), Global Server Load Balancer (GSLB), and methods for their operation in data networks are disclosed. The methods for load balancing may include receiving a query concerning a host name from a client, determining that there are two or more host servers associated with the host name, measuring various metrics associated with each of the two or more host servers and a local Domain Name Server (DNS), and based at least in part on the measurement, selecting a host server among the two or more host servers. The load balancing may also be based on a measured round trip time.

Application delivery controller and global server load balancer

Example embodiments of the present invention provide a method, an apparatus and a computer-program product for bandwidth management in a metro-cluster data replication environment. The method includes receiving an I/O at a first site and replicating the I/O from the first site to a second site by synchronous replication. Load balancing is then determined between the first site and the second site. Respective portions of the I/O then may be replicated from at least one of the first site and the second site to a third site by asynchronous replication according to the load balancing.

Method and apparatus for bandwidth management in a metro cluster environment

Generally described, the present disclosure is directed to an eventually consistent replicated data store that uses, for its underlying storage, a computer software library that provides a high-performance embedded database for data. The replicated data store employs a plurality of hosts interconnected to one another, allowing for writes to any host and full awareness of membership across all hosts. With the data replication framework disclosed herein, various modes are allowed to be built up on top of the core system.

Data replication framework

Embodiments of a distributed caching system are disclosed that cache data across multiple computing devices on a network. In one embodiment, a first cache system serves as a caching front-end to a distributed cluster of additional cache systems. The first cache system can distribute cache requests to the additional cache systems. The first distributed caching system can also serve as a cache server itself, by storing data on its own internal cache. For example, the first cache system can first attempt to find a requested data item on the internal cache, but, if the lookup results in a cache miss, the first cache system can search the additional cache systems for the data. In some embodiments, the first cache system is configured to identify additional requests for the first data item and duplicate the requested data item in order to respond to the additional requests, which allows for greater responsiveness to requests.

Distributed caching system

Processes are disclosed for embodiments of a caching system to utilize a snapshot file or other limited size data structure to store a portion of the data stored in a cache. The snapshot file can be stored on persistent or otherwise non-transitory storage so that, even in case of a restart, crash or power loss event, the data stored in the snapshot file persists and can be used by the caching system after starting up. The snapshot file can then be used to restore at least some data into the cache in cases where the cached data in the cache is lost. For example, in cases of a cold-start or restart, the caching system can load data from the snapshot file into the empty cache. This can increase the number of cache hits since the cache is repopulated with useful data at startup.

Cached data replication for cache recovery

Management of a data cache having a compressed portion and an uncompressed portion by adaptively and dynamically allocating the relative amount space each portion receives. The relative sizes are defined based on one or more cost metrics and benefit metrics. The metrics are selected based on the performance of an application utilizing the cache. An optimized benefit relative to the cost is defined. Application operations on the cache are sampled and the relative cost and benefit is determined for different ratios of uncompressed to compressed cache sizes. The size of the uncompressed portion relative to the compressed portion is then re-adjusted based on an optimal cost-to-benefit ratio for the application.

Dynamic selective cache compression

Processes and systems are disclosed for leasing a consumer system access to a producer system. The consumer system can communicate with a set of leasing agents to obtain the identity of a number of producer systems capable of providing the consumer system with access to a service. Each leasing agent may provide the consumer system with access to a producer system for a period of time. If the consumer system requires further access to a particular producer system, the consumer system can request renewal of the lease from the leasing agent that provided the initial lease of the producer system to the consumer system.

Vishal Parakh

Papers

Data replication framework

Distributed caching system

Cached data replication for cache recovery

Dynamic selective cache compression

Managing client access to a plurality of computing systems