Abstract:
Provided herein is a filing system creating and deleting apparatus, comprising an available inode pool in which inodes are to be stored, a file system object creating processor creating an inode allocating request signal in response to a file system creating request from an external resource, an inode allocating processor allocating at least one of inodes stored in the available inode pool in response to the inode allocating request signal, a file system object deleting processor creating an inode deallocating request signal in response to a file system deleting request from the external resource, an inode deleting thread collecting inodes requested to be deleted corresponding to the file system deleting request, and an inode deallocating processor deallocating the inode requested to be deleted from the inode deleting thread, wherein a number of inodes stored in the available inode pool is limited to no more than a maximum inode number in the available inode pool, and the inode requested to be deleted creates and deletes a file system stored in the available inode pool when the number of inodes stored in the available inode pool is smaller than the maximum inode number in the available inode pool.
Abstract:
The present invention relates to operating a parity chunk, and a parity chunk operating method, in an environment dividing data of a file by a chunk unit, constituting a stripe by calculating a parity for a failure recovery, dispersely storing a data chunk and a parity chunk included in the stripe in multiple data servers, comprising the steps of: storing a data chunk and a parity chunk of a stripe structure in data servers; generating a partial parity using a previous data chunk in the storing step of a update data chunk by the data server storing the data chunk; and transmitting a partial parity to the data server storing the parity chunk, and a data server apparatus for supporting the same are disclosed.
Abstract:
An exemplary embodiment provides an image processing apparatus, including: a first detector configured to detect a privacy image from a first image when the first image is input; a protection processor configured to scramble the privacy image to a privacy protection image using a scrambling key when the privacy image is detected; and an encryption processor configured to create a second image in which the first image including the privacy protection image is encrypted using an encryption key.
Abstract:
An exemplary embodiment of the present invention provides a flow switch, and an operating method thereof, and more particularly, to a flow switch for providing network virtualization, in which a quality of service (QoS) for a packet for each virtual machine is easily secured through flow information extracted from each of a plurality of packets when the plurality of packets generated by a plurality of virtual machines is input in a server virtualization environment, and an operating method thereof.
Abstract:
Provided herein is a distributed file system including a meta data server configured to store and manage meta data of a file; and a data server configured to store and manage data and attribute information of the file, wherein, in response to a writing request from a client, the meta data server stores data requested to be changed in a delta chunk of a read-only snapshot file, and connects the original file and the read-only snapshot file in an inode structure.
Abstract:
Provided is a device and a method for providing network virtualization, in which a method of dynamically mapping a processor includes extracting tenant information on a tenant and information on a virtual machine (VM) generated by a Cloud OS or controller; classifying virtual machine queues (VMQs) and processors to process the VMQs by tenant, and dynamically mapping the VMQs onto the processors by tenant.
Abstract:
An apparatus and a method for mapping of a tenant based dynamic processor, which classify virtual machine multi queues and processors processing the corresponding multi queues for each tenant that one or more virtual machines are belonged to, dynamically map the virtual machine multi queues that belong to the corresponding tenant to multi processors that belong to the corresponding tenant based on total usages of a network and a processor for each tenant to provide network virtualization to assure network traffic processing of virtual machines that belong to the same tenant not to be influenced by congestion of network traffic that belongs to another tenant.