The copy function of ETERNUS Disk storage system. A generic term for EC (Equivalent Copy) and OPC (One Point Copy).
For details, see the ETERNUS Disk storage system manual.
A volume that is directly or indirectly related to the master or the proxy. The alternative volume of a master volume is the proxy volume. The alternative volume of a proxy volume is the master volume or another proxy volume related to the master volume.
To add the detached slice to mirror volume making it accessible again as a component of the mirror volume.
The capacity on an SDX disk that is available for creating volumes. It equals the physical disk size rounded to a cylinder boundary, minus the private slice size.
The actual copying that takes place between master and proxy, after virtual completion of creating a snapshot or restoring the master when performing instant snapshot or instant restoring. When performing normal copy, the volume to which the data is being copied cannot be activated. However, when performing background copy, the volume to which the data is being copied can be activated for users to access the valid data.
To output the configuration of objects within a class to a configuration file in configuration table format.
A unit for disk capacity or data transmission. Usually, one block is 512 bytes.
Sector is used as a synonym.
One of the disk device identifiers managed by the udev function of OS. A file name of the by-id file is not changed even if a device name is changed. It is the symbolic link file corresponded to the device file in the /dev directory, and exists in the /dev/disk/by-id directory.
A resource categorized as a userApplication used to group resources into a logical collection.
For details, see the PRIMECLUSTER manual.
A component contained in PRIMECLUSTER responsible for detecting various abnormalities and state transition.
For details, see the PRIMECLUSTER manual.
A highly available and reliable system with multiple nodes connected via a special network.
For details, see the PRIMECLUSTER manual.
To connect more than one physical disk. It is helpful when using multiple disks as one virtual mass disk.
A group whose type attribute is "concat." It is also the set of disks that will be concatenated.
Database storing the configuration data of the objects managed by GDS. The configuration database is stored on a private slice.
To add an SDX disk to a disk group. User can also connect a keep disk, a single disk, or an undefined disk. Spare disk will be connected automatically by GDS when required.
To add a group to another group.
The former group is called a "lower level group," and the latter, a "higher level group."
For example, when more than one stripe group is connected to a mirror group that is the highest level group, the mirror volume that is created within the highest level group will be a logical device unit that will be both mirrored and striped.
To change a configuration table according to a restoration environment in order to restore the object configuration of the class in the environment that is different from the original.
The master volume or the proxy volume that is described as a file system or a swap area in the /etc/fstab file.
To temporarily detach a mirror slice from a mirror volume, making it inaccessible as a component of the mirror volume.
Other than by user command, GDS will automatically detach slices to maintain synchronization when an I/O error of a mirrored slice is detected.
For Linux specification, device names are changed by changing recognized order of devices in starting the system. Device name changes of disks may occur when restarting the system after occurring a disk error or adding and removing a disk.
A disk unit which combines multiple physical disks to be handled as one virtual disk. Spreading the access across multiple physical disks greatly improves performance and reliability. There are several access methods which are referred to as RAID (Redundant Array of Inexpensive Disks) levels. Main RAID level includes RAID 0 (striping), RAID 1 (mirroring), and RAID 5 (striping with parity).
A collection of SDX objects. Disk classes may be called "classes." Physical disks can be registered with a class and accordingly are managed by GDS. Classes are classified into root classes, local classes, and shared classes.
A collection of SDX disks and lower level groups. Disk groups may be called "groups." SDX disks and lower groups that belong to the same group become the unit of mirroring, striping, concatenating, or switching according to the type attribute of the group (mirror, stripe, concatenation, switch).
One cluster system built by use of PRIMECLUSTER.
A server (computer) that does not belong to a cluster system is also referred to as a domain.
An abbreviation for Equivalent Copy function. It is one of the Advanced Copy functions provided by ETERNUS Disk storage system. It keeps duplicating disk data within a disk unit. By performing detaching control with software, a replicate of the disk will be created.
For details, see the ETERNUS Disk storage system manual.
EC request from the host (server) to ETERNUS Disk storage system.
For details, see the ETERNUS Disk storage system manual.
A virtual device that provides multiple access paths with the emcpower driver. Being able to use multiple paths enables spreading out of the load when accessing the disk, and increases the availability at the time of path failure.
Software which allows various operations including GDS configuration setting and operation management using the Web screen.
Physical properties of a disk such as the numbers and sizes of cylinders, heads, and sectors.
Shared file system that features simultaneous accessibility that maintains consistency between multiple nodes to which a shared disk unit is connected and enables the continuation of processing on another node even if a certain node fails.
For details, see "PRIMECLUSTER Global File Services Configuration and Administration Guide."
A software that manages shared file system.
For details, see "PRIMECLUSTER Global File Services Configuration and Administration Guide."
To copy disk data created within the device using the disk array's hardware function so that no load is imposed on the server or the SAN.
A feature realizing automatic mirroring recovery while maintaining data accessibility at times of mirror disk failure.
A feature allowing users to exchange the faulty physical disk components without stopping the application in use. Also referred to as active exchange.
A standby slice that will be the operating slice when the active disk of the switch group is changed through a disk switch. The standby slice is a gathering of physically continuous blocks in the standby disk that belongs to the switch group."
To instantly restore the master's data, by overwriting the master using the parted proxy data. Instant restore can be created by using the -e instant option with the sdxproxy RejoinRestore command or the -e instant option with the sdxproxy Restore command.
To create a snapshot of a master instantly by cooperation with the OPC (One Point Copy) function provided by ETERNUS Disk storage system. Also it is a created snapshot. Snapshots can be created by using the -e instant option with the sdxproxy Part command.
To relate a pair of volumes or a pair of groups as a master and a proxy and synchronize the pair.
A high-speed resynchronization copying process. Available when just resynchronization mode is valid.
A disk that retains the disk format and data when being registered with a class or being connected to a group.
A "local" type class. Objects that belong to a local class can be used only on the current node. With the local class, disks connected to the current node, excluding system disks, can be registered.
The component of a logical volume. A logical slice consists of contiguous blocks of one physical disk or multiple blocks of multiple physical disks. A mirror volume consists of one or more logical slices that are mirrored. A single volume, a stripe volume, a volume in a concatenation group, and a switch volume consist of a single logical slice. Logical slices may be called "slices."
A logical disk device that provides the user with a block device interface as well as a physical slice (disk partition). The user can access data on a physical disk through accessing the logical volume. Logical volumes are classified into single volumes, mirror volumes, stripe volumes, volumes in concatenation groups, and switch volumes. Logical volumes may be called "volumes."
A group that belongs to another group. A volume cannot be created within the lower level group.
A shadow group that belongs to another shadow group. A shadow volume cannot be created in the lower level shadow group.
The original group from which the proxy group is copied.
The term "master" can refer to both master group and master volume.
The original volume from which the proxy volume is copied.
Volumes within master group are also called master volumes.
The term "master" can refer to both master volume and master group.
A group whose type attribute is "mirror." It is the set of disks and lower level groups, which will be mirrored each other.
A logical slice that composes a mirror volume. A mirror volume is created by mirroring more than one logical slice. For each mirror volume, there will be one logical slice in each disk and each lower level group belonging to the mirror group.
A volume that is created within a mirror group. Its data will be made redundant by mirroring.
To maintain redundancy by allocating the same data to multiple slices. A feature which allows you to continue access to a volume if one or more valid slices exist in the volume, even if some failures occur on the other slices.
To mirror the local disks (such as an internal disk) of more than one server via the network.
A virtual device that provides multiple access paths with the device-mapper driver which is the standard OS multipath function of RHEL. Being able to use multiple paths enables spreading out of the load when accessing the disk, and increases the availability at the time of path failure.
A virtual device that provides multiple access paths with the mphd driver. The user will be able to continue operation, even if an access path failure occurs.
A virtual device that provides multiple access paths with the mplb driver. Being able to use multiple paths enables spreading out of the load when accessing the disk, and increases the availability at the time of path failure.
A group whose type is "netmirror." It is the set of disks, which will be mirrored among servers.
A logical slice that composes a netmirror volume. A netmirror volume is created by mirroring more than one netmirror slice. For each netmirror volume, there will be one logical slice in each disk belonging to the netmirror group.
A volume that is created within a netmirror group. Its data will be made redundant by mirroring among servers.
An abbreviation for Non-Volatile Update Region Map. This is information of the updated volume region stored on a private slice for just resynchronization.
An SDX object. Objects are the generic term for classes, groups, disks, volumes, and slices.
The generic term for SDX objects and shadow objects.
To change from one disk to another without stopping applications accessing the disk data.
An abbreviation for One Point Copy function. It is one of the Advanced Copy function provided by ETERNUS Disk storage system. Disk data within the disk unit at a certain point will be instantly copied to separate area in the disk unit. The logical copying process is completed immediately after the request to start copying is sent from the server. The actual copying process will be performed later in the background.
For details, see the ETERNUS Disk storage system manual.
OPC request from the host (server) to the ETERNUS Disk storage system.
For details, see the ETERNUS Disk storage system manual.
To temporarily separate proxy from master so that it may be accessed independently. The parted proxy becomes a snapshot with master data of the moment it was parted. Although parted, the master-proxy relationship is maintained.
The status of a proxy isolated from the master. When a proxy volume is parted, access to the proxy that is independent of access to the master is available.
A storage media possible of handling random input and output of data. Mainly used to refer to a set of physical slices.
A part of physical disk and a series of physically continued blocks. A disk partition. The starting sector number and the number of sectors occupied by the physical slice is registered with the disk label (partition table).
A cluster system configured using PRIMECLUSTER. For details, see the PRIMECLUSTER manual.
A special slice used by GDS for control. Part of the SDX disk. Cannot be used as a logical slice.
Replication group of master group.
It has the same volume configuration as the master group, and each volume is the proxy volume of the master volume that is within the master group.
The term "proxy" can refer to both proxy group and proxy volume.
A replication volume of a master volume. This may be called an alternative volume.
Proxy volumes and proxy groups may be called "proxies" collectively.
One of the OPC functions of ETERNUS Disk storage system. All data is copied at the first copy processing and only updated data (difference) is copied from the second time.
For details, see the manual for ETERNUS Disk storage system.
QuickOPC request that is from a host (server) to an ETERNUS Disk storage system.
For details, see the manuarul for ETERNUS Disk storage system.
An abbreviation Remote Equivalent Copy function. It is one of the Advanced Copy function provided by ETERNUS Disk storage system. REC is EC function available between ETERNUS Disk storage system boxes. For details, see the ETERNUS Disk storage system manual.
To add a physical disk to a certain class.
The registered disk is managed by GDS subsequently.
To relate a pair of volumes or a pair of groups as a master and a proxy and leave them parted.
The concept describing the elements managed by PRIMECLUSTER among hardware and software that compose the PRIMECLUSTER system. In a PRIMECLUSTER system, shared disk units and GDS disk classes are managed as resources.
For details, see the PRIMECLUSTER manual.
The database that manages resource information in a PRIMECLUSTER system. The resource database is managed by the PRIMECLUSTER CF cluster resource management facility. For details, see the PRIMECLUSTER manual.
The function that registers hardware units with the resource database in a PRIMECLUSTER system. This function is also referred to as automatic resource registration. To perform resource registration, use the PRIMECLUSTER CF clautoconfig command. For details, see "PRIMECLUSTER Cluster Foundation (CF) Configuration and Administration Guide" and "6.3 Shared Disk Resource Registration."
To create object configuration of a class according to the configuration table backed up to a configuration file.
To copy data between mirror slices to recover mirroring state when rebooting after a system failure, after a cluster failover, or after reattaching a detached slice to a volume.
Or, to copy data between master and proxy to match their data when rejoining a parted proxy to a master.
A "root" type class. Objects that belong to a root class can be used only on the current node. With the root class, system disks connected to the current node, disks to mirror with the system disk, spare disks, and disks on which snapshots of the system disks are created, can be registered.
For the environment conditions where a root class can be used, see "3.4 Managing System Disks."
An abbreviation for "Storage Area Network." It is a network connecting multiple servers and storage devices. Usually, Fibre Channel is deployed for connection.
The function that boots the OS from a disk array unit connected to the SAN instead of an internal disk.
One of the class attributes. The scope indicates a group of nodes on which objects of the class are available.
A disk registered with a disk class. SDX disks may be called single disks, keep disks, spare disks, and undefined disks according to their usage. SDX disks may be called "disks."
A virtual resource managed by GDS. SDX objects are the generic term for classes, groups, disks, volumes, and slices. SDX objects may be called "objects."
A collection of shadow objects. Disk classes and shadow classes may be called "classes" collectively. Physical disks managed as SDX disks in another domain and those as the hard copy destinations of SDX disk data copied with disk unit copy functions can be registered with a shadow class and accordingly can be managed and accessed as shadow objects. Objects that belong to a shadow class can be used only on the current node.
A disk registered with a shadow class. SDX disks and shadow disks may be called "disks" collectively.
A group created in a shadow class. The shadow group is a collection of shadow disks and lower shadow groups. Disk groups and shadow groups may be called "groups" collectively.
A virtual resource managed by GDS. Shadow objects are the generic term for shadow classes, shadow groups, shadow disks, shadow volumes, and shadow slices. These shadow objects are available on a node where GDS Snapshot has been implemented. SDX objects and shadow objects may be called "objects" collectively.
The slice that composes a shadow volume. Logical slices and shadow slices may be called "slices" collectively.
A volume created in a shadow group. Logical volumes and shadow volumes may be called "volumes" collectively.
A "shared" type class. Objects that belong to a shared class are available on multiple nodes. With the shared class, disks connected to multiple nodes, including the current node, can be registered. Shared classes are also the units of resources used by cluster applications.
A file that declares physical connection configuration of nodes and shared disk units in a PRIMECLUSTER system. Specify the full path to a shared disk definition file with the -f option when executing the clautoconfig(8) command, and disk unit information in the shared disk definition file is registered with the PRIMECLUSTER resource database. For details, see "PRIMECLUSTER Cluster Foundation (CF) Configuration and Administration Guide" and "6.3 Shared Disk Resource Registration."
An SDX disk which does not belong to any group and is capable of creating single volumes.
A logical slice that consists a single volume. Single volume consists of one single slice. It is also a series of physically continued block in a single disk.
A volume that is created within a single disk that does not belong to a group. Its data is not made redundant.
To create a replica of a volume at a certain moment in another. It is also a created replica. GDS provides snapshot functions by slice detachment and by proxy volumes. Through use of proxy volumes, snapshot by synchronization and instant snapshot by OPC are available.
To create a snapshot of a volume at a moment to a slice by detaching a slice from a mirror volume temporarily. Also it is a created snapshot.
To create a snapshot of a master at a moment to a proxy by parting the master and proxy which were previously joined and synchronized. Also it is a created snapshot.
A reserved disk to recover mirroring when a mirrored disk fails.
When an I/O error is detected in a disk belonging to a mirror group or its lower level group, this spare disk will automatically be connected to recover normal mirroring.
Dell EMC's software. An abbreviation for Symmetrix Remote Data Facility. Using Symmetrix hardware functions, disk data is copied between Symmetrix cases. For details see the SRDF manual.
Disks and lower level groups by which striping is conducted. In other words, disks and lower level groups belonging to a stripe group.
A disk group whose type attribute is "stripe." This is a group of disks and lower level groups by which striping is conducted.
A volume that is created within a stripe group. Stripe volume can distribute the I/O load among multiple disks with its striping function. Its data will not be redundant.
To divide data into a certain size and then allocate the divided data alternately to multiple slices. Striping distributes the I/O load across multiple physical disks.
A group of the switch type. This is a gathering of an active disk and an inactive disk that are switched.
An operating logical slice comprising a switch volume. A switch volume consists of one switch slice. The switch slice is a gathering of physically continuous blocks in the active disk that belongs to the switch group.
A volume created in a switch group. With the disk switch function, the roles of the active and inactive disks can be switched. Data will not be redundant.
The state when mirroring is operating normally.
Or, a state when the master and proxy data are the same.
To copy data from a volume to an attached slice in order to make a mirrored state or to increase the multiplicity of the mirroring configuration, or to copy data between a master and a proxy in order to make their contents equivalent.
The disk that contains the Linux operating system running on the system. It refers to the entire disk containing slice operating as one of the following file systems (or a swap area).
/, /usr, /var, /boot, /boot/efi, or a swap area
The volume currently running as one of the following file systems (or a swap area).
/, /usr, /var, /boot, /boot/efi, or a swap area
A group that does not belong to any other group. Volumes can be created within the highest level group.
the highest level shadow group
A shadow group that does not belong to another shadow group. A shadow volume can be created in the highest level shadow group.
Dell EMC's software. Using Symmetrix hardware functions, disk data is copied within a Symmetrix case. For details see the TimeFinder manual.
One of the statuses of QuickOPC session. The status while updated parts are being recorded.
For details, see the manual for ETERNUS Disk storage system.
An SDX disk that is neither a single disk, keep disk or a spare disk. Its usage is not specified yet.
To copy (overwrite) data from a master to a parted proxy. The updated proxy is a snapshot that retains an image of the master at a certain moment.
