Hardware architecture
The basic system architecture of a VSP G1000, VSP G1500, and VSP F1500 single controller includes virtual storage directors (microprocessors) are shared across the cache, front-end directors (host adapters), and back-end directors (disk adapters), providing processing power where and when it is needed, without latency or interruption.
This architecture significantly increases the I/O throughput, up to three times the speed of the VSP system. The system provides a highly granular upgrade path, allowing the addition of drives to the drive chassis, and components such as virtual storage directors to the controller chassis as storage needs increase.
VSP G1000, VSP G1500, and VSP F1500 RAID implementation
The benefits of RAID technology are low cost, high reliability, and high I/O performance of data storage devices. To gain these benefits, this storage system supports RAID levels 1, 5, and 6.
Array groups and RAID levels
An array group (also called parity group) is the basic unit of storage capacity for the storage system. In RAID 1, an array group is a set of four physical drives where one drive is installed in the same location in each of four contiguous disk trays in the same drive chassis. The following figure shows part of an SFF drive chassis where two RAID 1 array groups are set up. RAID 1 is shown by yellow rectangles. The RAID 2 is shown by the red rectangles.
The storage system supports the following RAID levels: RAID 1, RAID 5, and RAID 6. When configured in four-drive RAID 5 parity groups (3D+1P), 75% of the raw capacity is available to store user data, and 25% of the raw capacity is used for parity data.
RAID 1
The following two figures illustrate RAID 1 configurations. The tables following the figures describe each configuration.
|
Item |
Description |
|
Description |
Mirror disks (duplicated writes). Two disk drives, plus primary and secondary disk drives, compose a RAID pair (mirroring pair) and the identical data is written to the primary and secondary disk drives. The data is distributed on the two RAID pairs. |
|
Advantage |
RAID 1 is highly usable and reliable because of the duplicated data. It has higher performance than ordinary RAID 1 (when it consists of two disk drives) because it consists of the two RAID pairs. |
|
Disadvantage |
Requires disk capacity twice as large as the user data. |
|
Item |
Description |
|
Description |
Mirror disks (duplicated writing). The two parity groups of RAID 1(2D+2D) are concatenated and data is distributed on them. In the each RAID pair, data is written in duplicate. |
|
Advantage |
This configuration is highly usable and reliable because of the duplicated data. It has higher performance than the 2D+2D configuration because it consists of the four RAID pairs. |
|
Disadvantage |
Requires disk capacity twice as large as the user data. |
RAID 5
A RAID 5 array group consists of four or eight drives (3D+1P) or (7D+1P). The data is written across the four drives or eight drives in a stripe that has three or seven data chunks and one parity chunk. Each chunk contains either eight logical tracks (mainframe) or 768 logical blocks (open). This RAID 5 implementation minimizes the write penalty incurred by standard RAID 5 implementations by keeping write data in cache until the entire stripe can be built, and then writing the entire data stripe to the drives. The 7D+1P RAID 5 configuration increases usable capacity and improves performance.
The following two figures illustrate RAID 5 configurations. The tables following the figures describes each configuration.
|
Item |
Description |
|
Description |
Data is written to multiple disks successively in units of block (or blocks). Parity data is generated from data of multiple blocks and written to disk. |
|
Advantage |
RAID 5 supports transaction operations that mainly use small size random access because each disk can receive I/O instructions independently. It can provide high reliability and usability at a comparatively low cost by virtue of the parity data. |
|
Disadvantage |
Write penalty of RAID 5 is larger than of RAID 1 because pre-update data and pre-update parity data must be read internally as the parity data is updated when data is updated. |
|
Item |
Description |
|
Description |
Two or four parity groups (eight drives) are concatenated. The data is distributed and arranged in 16 drives or 32 drives. |
|
Advantage |
If a RAID 5 (3D+1P) parity group becomes a performance bottleneck, you might improve performance through the added drives in a RAID 5 (7D+1P) configuration. |
|
Disadvantage |
The impact when two drives are blocked is significant because twice or four times the numbers of LDEVs are arranged in the parity group when compared with RAID 5 (3D+1P). However, the chance that the read of the single block in the parity group cannot be performed due to failure is the same as that of RAID 5 (3D+1P). |
Figure 1: Sample RAID 5 3D + 1P Layout (Data Plus Parity Stripe) shows RAID 5 data stripes mapped across four physical drives. Data and parity are striped across each drive in the array group. The logical devices (LDEVs) are dispersed evenly in the array group, so that the performance of each LDEV within the array group is the same. This figure also shows the parity chunks that are the Exclusive OR (XOR) of the data chunks. The parity chunks and data chunks rotate after each stripe. The total data in each stripe is 2304 blocks (768 blocks per chunk) for Open-systems data. Each of these array groups can be configured as either 3390-x or OPEN-x logical devices. All LDEVs in the array group must be the same format (3390-x or OPEN-x). For Open systems, each LDEV is mapped to a SCSI address, so that it has a track identifier (TID) and logical unit number (LUN).
RAID 6
A RAID 6 array group consists of eight drives (6D+2P). The data is written across the eight drives in a stripe that has six data chunks and two parity chunks. Each chunk contains 768 logical blocks.
In RAID 6, data can be assured when up to two drives in an array group fail. Therefore, RAID 6 is the most reliable of the RAID levels.
The following figure illustrates the RAID 6 configuration and the table describes the configuration.
|
Item |
Description |
|
Description |
Data blocks are scattered to multiple disks in the same way as RAID 5 and two parity disks, P and Q, are set in each row. Therefore, data can be assured even when failures occur in up to two disk drives in a parity group. |
|
Advantage |
RAID 6 is much more reliable than RAID 1 and RAID 5 because it can restore data even when failures occur in up to two disks in a parity group. |
|
Disadvantage |
The parity data P and Q must be updated when data is updated, RAID 6 imposes a write heavier than that on RAID 5. Performance of the random writing is lower than RAID 5 when the number of drives makes a bottleneck. |
LDEV striping across array groups
In addition to the conventional concatenation of RAID 1 array groups (4D+4D), the storage system supports LDEV striping across multiple RAID 5 array groups for improved logical unit performance in open-system environments. The advantages of LDEV striping are:
- Improved performance, especially of an individual logical unit, due to an increase in the number of drives that constitute an array group.
- Superior workload distribution: If the workload of one array group is higher than another array group, you can distribute the workload by combining the array groups, thereby reducing the total workload concentrated on each specific array group.
The supported LDEV striping configurations are:
- LDEV striping across two RAID 5 (7D+1P) array groups. The maximum number of LDEVs in this configuration is 1000. See the following figure.
- LDEV striping across four RAID 5 (7D+1P) array groups. The maximum number of LDEVs in this configuration is 2000. See Figure 2: LDEV striping across four RAID 5 (7D+1P) array groups.
All drives and device emulation types are supported for LDEV striping. LDEV striping can be used with all storage system data management functions.
Control unit images, logical volume images, and logical units
This section provides information about control unit images (CU), logical volume images (LVI), and logical units (LU).
CU images
The storage system is configured with one control unit image for each 256 devices (one SSID per 64 LDEVs or 256 LDEVs) and supports a maximum of 255 CU images in the primary logical disk controller (LDKC).
The storage system supports the control unit (CU) emulation type 2107.
The mainframe data management features of the storage system can restrict CU image compatibility.
For more information on CU image support, see the Mainframe Host Attachment and Operations Guide, or contact your Hitachi Vantara account team.
Logical volume images
The VSP G1000, VSP G1500, and VSP F1500 storage systems support the following Mainframe LVI types:
- 3390-3
- 3390-3R
- 3390-9
- 3390-L
- 3390-M
The LVI configuration of the storage system depends on the RAID implementation and physical data drive capacities. To access the LDEVs, combine the logical disk controller number (00), CU number (00-FE), and device number (00-FF). All control unit images can support an installed LVI range of 00 to FF.
For maximum flexibility in LVI configuration, the storage system provides the Virtual LVI feature. Using Virtual LVI, users can configure multiple LVIs under a single LDEV. For further information on Virtual LVI, see the Provisioning Guide for Mainframe Systems.
Logical units
The storage system is configured with OPEN-V logical unit types. The OPEN-V logical unit size can vary from 48.1 MB to 4 TB. For information about other logical unit types (for example, OPEN-9), contact Hitachi Vantara support.
For maximum flexibility in LU configuration, the storage system provides the Virtual LUN feature. Using Virtual LUN, users can configure multiple LUs under a single LDEV. For further information on Virtual LUN, see the Provisioning Guide for Open Systems.
Mainframe operations
This section provides high-level descriptions of mainframe compatibility, support, and configurations.
Mainframe compatibility and functionality
In addition to full System Managed Storage (SMS) compatibility, the storage system provides the following functions and support in a Mainframe environment:
- Sequential data striping
- Cache fast write (CFW) and DASD fast write (DFW)
- Enhanced dynamic cache management
- Extended count key data (ECKD) commands
- Multiple Allegiance
- Concurrent Copy (CC)
- Peer-to-Peer Remote Copy (PPRC)
- FlashCopy®
- Parallel Access Volume (PAV)
- Hyper Parallel Access Volume (HPAV)
- Priority I/O queuing
- Red Hat Linux for IBM® S/390® and IBM® zSeries®
- SUSE Linux for IBM® S/390® and IBM® zSeries®
- zHyperWrite™ for DB2® (for details, see Hitachi Virtual Storage Platform G1000, G1500, F1500 Hitachi TrueCopy® for Mainframe User Guide)
- zHPF Extended Distance II (VSP G1000, VSP G1500, and VSP F1500 microcode 80-05-0x or later)
- FICON® Dynamic Routing
- FICON® Forward Error Correction (VSP G1000, VSP G1500, and VSP F1500 microcode 80-05-0x or later)
Mainframe operating system support
The VSP G1000, VSP G1500, and VSP F1500 storage systems support most major IBM mainframe operating systems. For more information about supported operating systems, see the mainframe support matrix on https://support.hds.com/en_us/interoperability.html.
Mainframe configuration
After a storage system installation is complete, users can configure the storage system for mainframe operations.
See the following user documents for information and instructions about configuring your storage system for mainframe operations:
- The
Mainframe Host Attachment and Operations Guide, provides instructions for configuring the storage system for mainframe operations, including FICON attachment, hardware definition, cache operations, and device operations.
For detailed information about FICON connectivity, FICON or Open intermix configurations, and supported HBAs, switches, and directors for VSP G1000, VSP G1500, and VSP F1500, contact customer support.
- The System Administrator Guide provides instructions for installing, configuring, and using Device Manager - Storage Navigator to perform resource and data management operations on the storage systems.
- The Provisioning Guide for Mainframe Systems provides instructions for converting single volumes (LVIs) into multiple smaller volumes to improve data access performance.
Open-systems operations
This section provides high-level descriptions of open-systems compatibility, support, and configuration for storage systems.
Open-systems compatibility and functionality
The VSP G1000, VSP G1500, and VSP F1500 storage systems support many features and functions for the open-systems environment, including:
- Compatibility with most iSCSI adapters, Fibre Channel host bus adapters (HBAs), and Fibre Channel-over-Ethernet (FCoE) converged network adapters (CNAs)
- Multi-initiator I/O configurations with multiple host systems attached to the same Fibre Channel interface
- Fibre Channel Arbitrated Loop (FC-AL) topology, supported on ports up to 8 Gbps
- Fibre Channel fabric topology using direct attach point-to-point connections, supported on ports up to 16 Gbps
NoteWhen using 16-Gbps FC front-end director in an FC-AL topology, the transmission speed is limited to 8 Gbps.
- Command tag queuing
- Industry-standard failover and logical volume management software
- SNMP remote storage system management
The global cache feature in VSP G1000, VSP G1500, and VSP F1500 enables any port to have access to any logical unit in the storage system. Each logical unit can be assigned to multiple ports to provide I/O path failover and load balancing (with the appropriate middleware support, such as Hitachi Global Link Manager) without sacrificing cache coherency.
Users should plan for path failover (alternate pathing) to ensure the highest data availability. The logical units can be mapped for access from multiple ports or multiple target IDs. The number of connected hosts is limited only by the number of Fibre Channel ports installed and the requirement for alternate pathing within each host. If possible, the primary path and alternate paths should be attached to different channel cards.
Open-systems host platform support
The VSP G1000, VSP G1500, and VSP F1500 storage systems support most major open-system operating systems, including Windows®, Solaris, IBM AIX®, Linux, HP-UX, and VMware. For complete information about supported operating systems, visit https://support.hds.com/en_us/interoperability.html.
For a complete list of the storage system user guides, including the Open-Systems Host Attachment Guide, see the Product Overview.
System configuration
After installing the storage system is complete, you can configure the storage system for open-systems operations.
Refer to the following documents for information and instructions about configuring your storage system for open-systems operations:
- The
Open-Systems Host Attachment Guide provides information and instructions to configure the storage system and data storage devices for attachment to the open-systems hosts.
NoteThe storage system queue depth and other parameters are adjustable. See the Open-Systems Host Attachment Guide for queue depth and other requirements.
- The System Administrator Guide provides instructions for installing, configuring, and using Device Manager - Storage Navigator to perform resource and data management operations on the storage system.
- The
Provisioning Guide for Open Systems describes and provides instructions for configuring the storage system for host operations, including FC port configuration, LUN mapping, host groups, host modes and host mode options, and LUN security.
Each FC port on the storage system provides addressing capabilities for up to 2,048 LUNs across as many as 255 host groups, each with its own LUN 0, host mode, and host mode options. Multiple host groups are supported using LUN security.
- The Hitachi Alert Notification Guide describes the SNMP API interface for the storage systems and provides instructions for configuring and performing SNMP operations.
- The Provisioning Guide for Open Systems provides instructions for configuring multiple custom volumes (logical units) under single LDEVs on the VSP G1000, VSP G1500, and VSP F1500.
Host modes and host mode options
The VSP G1000, VSP G1500, and VSP F1500 storage systems connect multiple server hosts of different platforms to each of its ports.
When your system is configured, the hosts connected to each port are grouped by host group or by target. For example, if Solaris and Windows hosts are connected to separate iSCSI or FC ports, or through a switch to a single iSCSI or FC port, a host group is created for the Solaris hosts and another host group is created for the Windows hosts. The appropriate host mode and host mode options are assigned to each host group. The host modes and host mode options enhance compatibility with supported platforms and environments.
Use Device Manager - Storage Navigator to configure host groups, host modes, and host mode options. For more information about host groups, host modes, and host mode options, see the System Administrator Guide.
Device Manager - Storage Navigator program
Device Manager - Storage Navigator is the GUI that accesses the features in the microcode.
The GUI is also used to set up and monitor the storage system. It can be installed on a PC, laptop, or workstation. It communicates via a LAN to the SVP in the storage system. The SVP obtains storage system configuration and status information and sends user-initiated commands to the storage system. Device Manager - Storage Navigator displays detailed storage system information and allows you to configure and perform storage operations on the system.
Device Manager - Storage Navigator is a Java® applet program that can run on any machine that supports a Java Virtual Machine (JVM). A PC hosting the Device Manager - Storage Navigator software is called a remote console. Each time a remote console accesses and logs on to the SVP, the Device Manager - Storage Navigator applet is downloaded from the SVP to the remote console. The following figure shows the remote console and SVP configuration for Device Manager - Storage Navigator.
For more information about Device Manager - Storage Navigator, see the System Administrator Guide.
