Hitachi, Ltd.

Akiyoshi Tsuchiya

Tomohiro Kawaguchi

Tsukasa Shibayama

Hitoshi Kamei

Masanori Takata

Hitoshi Arai

Yoji Nakatani

Hiroto Ebara

Akira Yamamoto

Hideo Saito

Masakuni Agetsuma

Mitsuo Hayasaka

Takaki Nakamura

Masanori Takada

Takuya Oda

Atsushi Sutoh

Keiichi Matsuzawa

Jun Nemoto

Azusa Jin

Miho Kobayashi

Yuki Sakashita

Yasushi Miyata

Kazuhito Yokoi

Go Kojima

Misato Yoshida

Hiroyuki Osaki

Hieuhanh Le

Provided is a resource allocation determination method for a VM/container, volume, and the like created as a new VM/container or volume without exceeding an upper limit of a computer resource of a node in an HCI environment. In order to determine allocation of at least one of a virtual machine, a container, and a volume in a system of the HCI environment, a use state of a computer resource shared by a virtual machine and a storage controller operating on each node is managed, and an allocation destination node of the new virtual machine, container, or volume is determined based on the use state without exceeding an upper limit of a computer resource of the allocation destination node.

Virtual machine storage controller selection in hyperconverged infrastructure environment and storage system

Example implementation described herein are directed to a mechanism to provision data volume which requires remote data copy between separated clusters, especially for the container platform. For a request to create a volume made to the clusters, example implementations can involve creating a first volume in a first cluster; obtaining volume information of a corresponding second volume from a second cluster; configuring the first volume and the corresponding second volume in the second cluster to have a remote copy relationship based on the obtained volume information; and setting access from the container to the first volume and the corresponding second volume based on the remote copy relationship.

Dynamic volume provisioning for remote replication

Example implementations described herein involve systems and methods which automatically determine volumes to be replicated for disaster recovery based on the execution priority of an application which uses the volumes. Such example implementations can involve systems and methods involving creating a volume in a first storage system for each of one or more containers newly launched on one or more servers managing a container orchestrator; and establishing replication of the volume for the each of the newly launched one or more containers to a second storage system in order from highest container priority to lowest container priority.

Volume remote copy based on application priority

Example implementations described herein are directed to searching ML models to facilitate IO load prediction. Even if there is no existing ML model applicable to the underlying container, example implementations described herein obtain the ML model for the nearest container. Through such example implementations, the underlying systems can reduce the total ML models maintained for the prediction of IO load and can enhance ML models with enough learning data for each ML model.

Method of selecting a machine learning model for performance prediction based on versioning information

Example implementations described herein are directed to backup and restore of a public cloud and data center for disaster recovery. In a backup process of the disaster recovery, the VM root device information (a public cloud) and storage system volume information (a data center) in a main site are transferred to the disaster recovery tool. In a restoration process, the DR Tool specifies the storage system volume for backup to attach the VM in a backup site with the above information.

Disaster recovery collaborated with storage system and backup appliance in hybrid cloud environment

Each file storage apparatus of a plurality of file storage apparatuses stores a file system, and associates and stores paths of elements in the file system and archive destinations of the elements in an archive storage apparatus. When the file system is operated, each file storage apparatus transmits archive data of an element as an operation target, and operation information including operation details to the archive storage apparatus. The archive storage apparatus receives the archive data and the operation information, stores the archive data, and stores consistency information including the operation information and archive versions indicating a reception order of the operation information. A first file storage apparatus executes a synchronization process of acquiring the consistency information from the archive storage apparatus, correcting inconsistency between the acquired consistency information and consistency information including archive versions earlier than the acquired consistency information, and reflecting the file system on the consistency information.

Computer system, file storage controller, and data sharing method

Example implementations described herein are directed to Input/Output (I/O) path reservation with out of band management. In example implementations, for failure of a storage orchestrator to delete the path between the container and the first volume, the example implementations described herein are directed to deleting a path between the first volume and a quorum volume; and establishing an I/O path between the new container and the second volume.

I/O path reservation with out of band management

Virtual machine and volume allocation in hyperconverged infrastructure environment and storage system

Provided is a resource allocation determination method for a VM/container, volume, and the like created as a new VM/container or volume without exceeding an upper limit of a computer resource of a node in an HCl environment. In order to determine allocation of at least one of a virtual machine, a container, and a volume in a system of the HCl environment, a use state of a computer resource shared by a virtual machine and a storage controller operating on each node is managed, and an allocation destination node of the new virtual machine, container, or volume is determined based on the use state without exceeding an upper limit of a computer resource of the allocation destination node.

A highly accurate data migration pattern is decided on the bases of usages of a plurality of kinds of resources and by taking influence of execution of data migration on each resource into consideration. In a distributed storage system, a monitor acquires a per-period usage of each of the plurality of kinds of the resources per data, a performance simulator estimates a per-period usage of each resource on a migration destination node when the data is migrated by using the acquired per-data resource usage and calculates an estimate of a latency from the estimated usage of each resource and an optimizer decides a migration pattern of the data on the basis of the estimate of the latency.

Distributed storage system and data migration method

When a second storage device in a second site receives an update notification regarding a first element updated according to a write request from a first host, from a first storage device in a first site, for a second element specified on the basis of the update notification, the second storage device manages data of the first element corresponding to the second element as latest data. The first storage device provides a first volume capable of including a plurality of first elements. The second storage device provides a second volume capable of including a plurality of second elements corresponding to the plurality of first elements. When the second storage device receives a read request from a second host, the second storage device determines whether or not data of the first element corresponding to a read source second element (second element specified from the read request) is the latest data. When a determination result is true, the second storage device acquires the latest data from the first storage device and returns the latest data to the second host.

Storage system and data transfer control method

According to one embodiment, a storage system includes a representative node and a plurality of general nodes including a plurality of resources. Each of the general nodes stores resource status information indicating respective statuses of the plurality of resources in a first storage unit thereof. The representative node stores resource status information collected from the plurality of general nodes in a second storage unit thereof, decides whether to acquire the resource status information from the first storage unit of the general node or to acquire the resource status information from the second storage unit based on a received request, and transmits the resource status information acquired from a decided acquisition destination to an issuing source of the request.

Storage system and information management method having a plurality of representative nodes and a plurality of general nodes including a plurality of resources

A file server apparatus includes a second file system volume including cache data of a first file system volume stored in a shared file storage apparatus. The file server apparatus determines the operation type indicated by an access request to the first file system volume from a client. When the operation type is directory operation, the file server sends an instruction of directory operation to the shared file storage apparatus, and then transmits a completion response to the client. When the operation type is file operation, the file server apparatus executes a file operation in the second file system volume, transmits a completion response to the client, and then transmits a file operation instruction to the shared file storage apparatus.

File server apparatus having shared file system with file and directory operation processing

An active control program and a passive control program included in the same program cluster in a plurality of storage nodes are each arranged in the storage nodes different from each other. Each of the storage nodes is configured to include a plurality of active or passive control programs. When any one of the plurality of the passive control programs is switched to active, a change of an operation status is made for the different passive control program operating in the storage node that includes the passive control program switched to active.

Storage system and storage control method

A computer system for sharing objects comprises multiple front-end nodes and a back-end node for providing a namespace shared by the front-end nodes. A first front-end node that is any one of the front-end nodes manages local-lock counts for each of objects managed by the first front-end node. The local-lock count is the number of locks taken by operations to the objects including the directory and sub-objects thereof. Every time receiving a request for processing a target object from a client, the front-end node requests the back-end node to lock a parent directory of the target object if the parent directory has not been locked by the back-end node. The front-end node sends a response of the received request to the client regardless of whether or not the parent directory has been unlocked by the back-end node. If the local-lock count of the parent directory is zero, the first front-end node requests the back-end node to unlock the parent directory.

Growing community of inventors

Tokyo, Japan

Akiyoshi Tsuchiya