International Business Machines Corporation

Reshu Jain

Prasenjit Sarkar

Rui Zhang

Zhenhua Hu

Yonggang Hu

Karan Gupta

Himabindu Pucha

Aameek Singh

Eric Kevin Butler

Dinesh Kumar Subhraveti

Mihail Corneliu Constantinescu

Kavita Chavda

Mohit Saxena

Prashant Pandey

Mark James Seaman

Richard J Ayala, Jr

A method for multi-layer quality of service (QoS) management in a distributed computing environment includes receiving a workload and identifying a workload QoS class for the workload. The workload is scheduled to be run compute nodes. The workload QoS class is translated to a plurality of different resource layer QoS classes, and the plurality of QoS classes are communicated to a workload execution manager of the compute nodes. The workload execution manager extends the plurality of different resource layer QoS classes to respective resource layers involved in the running of the workload.

Multi-layer QOS management in a distributed computing environment

A system for multi-layer quality of service (QoS) management in a distributed computing environment includes a management node hosting a workload scheduler configured to receive a workload and identify a workload QoS class for the workload. The workload scheduler schedules running of the workload on compute nodes, translate the workload QoS class to a plurality of different resource layer QoS classes, and communicates the plurality of QoS classes to a workload execution manager of the compute nodes. The workload execution manager is configured to extend the plurality of different resource layer QoS classes to respective resource layers involved in the running of the workload.

Multi-layer QoS management in a distributed computing environment

A system for multi-layer quality of service (QoS) management in a distributed computing environment includes a management node hosting a workload scheduler operable to receive a workload and identify a workload QoS class for the workload. The system also includes a plurality of distributed compute nodes. A workload scheduler schedules running of the workload on the compute nodes, translates the workload QoS class to a storage level QoS class, and communicates the storage level QoS class to a workload execution manager of the compute nodes. The workload scheduler communicates the storage level QoS class to one or more storage managers where the storage managers manage storage resources, and the storage managers are operable to extend the storage level QoS class to the storage resources to support the workload QoS class.

A technique for multi-layer quality of service (QoS) management in a distributed computing environment includes receiving a workload to run in a distributed computing environment. A workload quality of service (QoS) class for the workload is identified, and the workload QoS class is translated to a storage level QoS class. The workload is scheduled to run on a compute node of the environment, and the storage level QoS class is communicated to a workload execution manager of the compute node. The storage level QoS class is communicated to one or more storage managers of the environment where the storage managers manage storage resources in the environment. The storage managers extend the storage level QoS class to the storage resources to support the workload QoS class.

A filesystem includes a pool of data blocks and a pool of index blocks. Non-overlapping subpools of data blocks, including a first subpool of data blocks and a second subpool of data blocks, are identified within the pool of data blocks. Likewise, non-overlapping subpools of index blocks, including a first subpool of index blocks and a second subpool of index blocks, are identified within the pool of index blocks. A first file group is mounted within a global namespace on a first mount point on the filesystem. A second file group is also mounted within the global namespace on a second mount point on the filesystem. The first subpool of data blocks and the first subpool of index blocks are allocated to the first mounted file group, and the second subpool of data blocks and the second subpool of index blocks are allocated to the second mounted file group.

Filesystem with isolated independent filesets

A system for multi-layer quality of service (QoS) management in a distributed computing environment includes a management node hosting a workload scheduler operable to receive a workload and identify a workload QoS class for the workload. The system also includes a plurality of distributed compute nodes. A workload scheduler is operable to schedule running of the workload on the compute nodes. The workload scheduler is operable to translate the workload QoS class to a storage level QoS class and communicate the storage level QoS class to a workload execution manager of the compute nodes. The workload scheduler communicates the storage level QoS class to one or more storage managers where the storage managers manage storage resources, and the storage managers are operable to extend the storage level QoS class to the storage resources to support the workload QoS class.

Embodiments of the present invention provide an approach for migrating virtual machines across network (e.g., WAN) separated data centers (e.g., storage clouds). Specifically, under embodiments of the present invention, a first storage system associated with a first data center is synchronized with a second storage system associated with a second data center via a storage system link. Then, a minimal state of a virtual machine is migrated from a first computer in the first data center to a second computer in the second data center via a WAN link. Using the minimal state, the virtual machine is stored in the second computer. Thereafter, the storage system link is terminated. In addition, as updated pages are received in memory of the first computer, they are migrated to the second computer via the WAN link. Once this migration is complete, the WAN link can be terminated. Therefore, embodiments of the present invention provide at least two forms of synchronization: computational synchronization and storage synchronization.

Migrating virtual machines across network separated data centers

A system for multi-layer quality of service (QoS) management in a distributed computing environment includes: a management node hosting a workload scheduler operable to receive a workload and identify a workload QoS class for the workload; and a plurality of distributed compute nodes, the workload scheduler operable to schedule running of the workload on the compute nodes. The workload scheduler is operable to: translate the workload QoS class to a storage level QoS class; communicate the storage level QoS class to a workload execution manager of the compute nodes; and communicate the storage level QoS class to one or more storage managers, the storage managers managing storage resources. The storage managers are operable to extend the storage level QoS class to the storage resources to support the workload QoS class.

A technique for multi-layer quality of service (QoS) management in a distributed computing environment includes: receiving a workload to run in a distributed computing environment; identifying a workload quality of service (QoS) class for the workload; translating the workload QoS class to a storage level QoS class; scheduling the workload to run on a compute node of the environment; communicating the storage level QoS class to a workload execution manager of the compute node; communicating the storage level QoS class to one or more storage managers of the environment, the storage managers managing storage resources in the environment; and extending, by the storage managers, the storage level QoS class to the storage resources to support the workload QoS class.

Embodiments of the invention relate to handling failures in a cluster of computer resources. The resources are represented as nodes in a dependency graph in which some nodes are articulation points and the removal of any articulation point due to a resource failure results in a disconnected graph. The embodiments perform a failover when a resource corresponding to an articulation point fails. The failover is to a local resource if the failed resource does not affect all local resources. The failover is to a remote resource if no local resource can meet all resource requirements of the failed resource, and to a remote resource running in a degraded mode if the remote resource cannot meet all of the requirements.

High-availability computer cluster with failover support based on a resource map

Embodiments of the invention relate to block layout and block allocation in a file system to support transparency of application processing. At least one copy of an application is replicated in a write affinity region of a secondary server, and at least one copy of the application is replicated in a wide striping region across a cluster file system. When the application is subject to failure, application processing is transferred from the failure location to the write affinity copy. At the same time, the failed application is rebuilt using the wide striping replication of the application. Once the application is rebuilt, processing may return to the failed location employing the rebuilt application.

Application recovery in a file system

Embodiments of the invention are directed to streaming virtual machine boot services over a network. An aspect of the invention includes booting a first virtual machine and recording data and metadata from a virtual machine boot image into a virtual machine boot file. The data and metadata are accessed in the process of booting the first virtual machine. The virtual machine boot image has setup information of the virtual machine type of the first virtual machine. The virtual machine boot file is configured for the virtual machine type of the first virtual machine. A descriptor is added to metadata of the virtual machine boot image, which references a location of the virtual machine boot file for the virtual machine type of the first virtual machine. When subsequently booting a second virtual machine of the same type of virtual machine as the first virtual machine, data are streamed from the virtual machine boot file to a virtual machine monitor of a second virtual machine without the need to stream data from the virtual machine boot image.

Streaming virtual machine boot services over a network

Embodiments of the invention relate to block layout and block allocation in a file system to support write transactions. Regions in a cluster file system are defined to support a block allocation, include both write affinity and wide striping region. An allocation map is maintained to define and support the block allocation regions. For each write transaction, a copy of data is written to both regions, and in the event there is a shortage of blocks in one of the regions, at least one block in the other region is dynamically converted to support the write transaction.

Growing community of inventors

Sunnyvale, CA, United States of America

Reshu Jain