International Business Machines Corporation

Eran Rom

Shmuel Ben-Yehuda

Avishay Traeger

Eran Borovik

Elliot Karl Kolodner

Sivan Tal

Sandeep R Patil

Michael Edward Factor

John Thomas Olson

Danny Harnik

Joseph S Glider

Duane Mark Baldwin

Dmitry Sotnikov

Paula Kim Ta-Shma

Dalit Naor

David Hadas

Smita J Raut

Demyn Lee Plantenberg

Nadav Yosef Har'el

Gil Vernik

Aner Hamama

Optimizing read operations to a storage device in a virtualized computing environment. A read system call is received by a virtual machine's (VM) operating system from an application running on the VM for data on a storage device. The VM is hosted by a hypervisor providing access to the storage device in a virtualized computing environment. A virtual file system (VFS) supported by the operating system is caused to call on a first para-virtualized file system (PVFS FE) supported by the operating system to execute the read operation. In response to determining by the PVS FE that the data to be read exists in a data cache of the PVFS FE, executing by the PVS FE the read operation from the cache of the PVFS FE.

Optimizing a file system interface in a virtualized computing environment

A method of and system for managing storage of an object for an object store is disclosed. The method and system may establish the object store. The object store may be configured to be used for a user for a workload-type for a set of objects. The method and system may provide, for the user for the workload-type, an object store quota. The method and system may provide, for the user for the workload-type, a meta-data quota. The method and system may process, consistent with both the object store quota and the meta-data quota, the set of objects.

Storage quota management

Machines, systems and methods of uploading data files, the method comprising a first client machine dividing a first file into N data chunks to be uploaded to a server, wherein the N data chunks are of size kX, where k is an integer and X is size of a minimal size data chunk, wherein X is known by the server and by at least a second client machine used for uploading a second file to the server in data chunks of size k'X; and uploading the first file to the server, wherein a first unique signature is calculated for the first file based on applying a signature function to a collection of signatures calculated for the minimal size data chunks of size X that make up the data chunks of size kX in the first file, wherein the uploading of the first file is accomplished by uploading the data chunks of size kX to the server in any order.

Efficient data deduplication in a data storage network

Systems and methods for optimizing write operations to a storage device in a virtualized computing environment comprise monitoring write operations issued by an application running on a virtual machine's (VM) operating system, wherein the VM is hosted by a hypervisor providing access to a storage device in a virtualized computing environment; and causing a virtual file system (VFS) supported by the operating system to call on a first para-virtualized file system (PVFS FE) supported by the operating system to execute a write operation, in response to determining that the write operation is to write data to the storage device, wherein data that is to be written to the storage device is first written to a VM memory area allocated to the VM and accessible to the hypervisor hosting the VM.

Systems and methods for associating a first process with a first state and a first computing environment initialized according to a first set of parameters, wherein a first task is to be performed under a first security context. The method further comprising associating a second process with a second state and a second computing environment initialized according to a second set of parameters; in response to the first process submitting a first request, the second process spawning a third process which has the second state; wherein the third process sets a security context for the third process to the first security context and the third process sets the computing environment for the third process according to a first a set of parameters; executing the third process under the first security context and in association with the second state; and executing the first task in the first computing environment.

Optimization of spawning request handling processes in a secured computing environment

Systems and methods for optimizing read operations from a storage device in a virtualized computing environment are provided. One method comprises monitoring read operations issued by an application running on a virtual machine's (VM) operating system, wherein the VM is hosted by a hypervisor providing access to a storage device in a virtualized computing environment; and causing a virtual file system (VFS) supported by the operating system to call on a first para-virtualized file system (PVFS FE) supported by the operating system to execute a read operation, in response to determining that the read operation is for reading target data from the storage device.

In accordance with one or more embodiments, an inode implemented file system may be utilized to support both offline and inline deduplication. When the first content is stored in the storage medium, one inode is used to associate a filename with the data blocks where the first content is stored. When a second content that is a duplicate of the first content is to be stored, then a parent inode is created to point to the data blocks in which a copy of the first content is stored. Further, two inodes are created, one representing the first content and the other representing the second content. Both inodes point to the same parent inode that points to the data blocks where the first content is stored.

Efficient file system object-based deduplication

A shared key and a private key unique to one or more resource sets are generated. Each of the resource sets identifies the resource. The shared key and the private key are transmitted to a security access point. The security access point controls access to the resource. A resource key is generated. The resource key is a result of a cryptographic function of the private key and a resource identifier. The resource key and the shared key are transmitted to one or more local security managers. Each of the local security managers is assigned to manage one of the resource sets. In accordance with one or more policies, the local security managers generate a credential using the resource key and the shared key. The credential is distributed to one or more authenticated clients. Further, the credential is used to grant the one or more authenticated clients access to the resource sets through the security access point.

Growing community of inventors

Tel Aviv, Israel

Eran Rom