Introducing the IBM DS8900F storage system
This chapter introduces the features and functions of the IBM DS8900F storage system that are available with IBM DS8000 Licensed Machine Code (LMC) 7.9.30 (bundle version 89.30.xx.x), which is referred to as Release 9.3. More information about functions and features is provided in subsequent chapters.
This chapter covers the following topics:
1.1 Introduction to the DS8900F system
The DS8000 family is a high-performance, high-capacity, highly secure, and resilient series of disk storage systems. The DS8900F family is the most advanced of the DS8000 offerings to date. All the DS8900F models are all-flash systems. The high availability (HA) multiplatform support, including IBM Z, and simplified management tools provide a cost-effective path to an on-demand world.
The IBM DS8900 family consists of the following high-performance models:
•IBM DS8980F: Analytic Class
•IBM DS8950F: Agility Class
•IBM DS8910F: Flexibility Class, rack-mounted
•IBM DS8910F: Flexibility Class, racked
The DS8900F models support the most demanding business applications with their exceptional all-around performance and data throughput. Some models are shown in Figure 1-1.
Figure 1-1 DS8900F frame (standard doors with side covers)
The DS8000 offers features that clients expect from a high-end storage system:
•High performance
•High capacity
•HA
•Security
•Cost efficiency
•Energy efficiency
•Scalability
•Business continuity and data protection functions
The DS8900F architecture is server-based. Two powerful POWER9 processor-based servers manage the cache to streamline drive I/Os, which maximizes performance and throughput. These capabilities are further enhanced with the availability of High-Performance Flash Enclosures (HPFEs). HPFE Gen2 is described in IBM DS8000 High-Performance Flash Enclosure Gen2 (DS8000 R9.0), REDP-5422.
The DS8900F is an all-flash storage system that is equipped with encryption-capable flash drives. High-density storage enclosures offer a considerable reduction in the footprint and energy consumption. Figure 1-2 shows the front view of a fully configured DS8950F frame.
Figure 1-2 DS8950F all-flash system with KVM console open
 
Reduced footprint: The DS8900F is housed in a 19-inch wide rack, and it has a smaller depth than its predecessor, which is the DS8880.
The DS8900F includes a power design that is based on intelligent Power Distribution Units (iPDUs), and non-volatile dual inline memory modules (NVDIMMs) to store data in case of a power outage. The iPDUs allow the DS8900F to achieve the highest energy efficiency in the DS8000 series. The DS8900F complies with ENERGY STAR specifications. All models are available for both three-phase and single-phase power attachment.
I/O enclosures are attached to the POWER9 processor-based servers with Peripheral Component Interconnect Express (PCIe) Generation 3 cables. The I/O enclosure has six PCIe adapter slots, and two zHyperLink ports.
The rack models also have an integrated keyboard and display that can be accessed from the front of the rack. A pair of small form-factor (SFF) Hardware Management Consoles (HMCs) are installed in the base rack management enclosure. The height of the rack is 40U for all units.
The DS8910F Rack-Mounted model comes without its own rack. It can attach to both mainframe or to distributed systems. When placed in a customer-supplied rack, a keyboard and display can be ordered. When integrated into the IBM 3907 z15™ T02 or LinuxONE LR1 business-class mainframe models, the DS8910F Rack-Mounted model shares the rack and the management keyboard and display.
 
Note: The DS8910F and its specific features are described in the IBM Power Systems Enterprise AI Solutions, REDP-5556.
Figure 1-3 on page 7 shows the various components within the base frame. The DS8950F and DS8980F expansion frame enables clients to add more capacity to their storage systems in the same footprint.
Figure 1-3 DS8950F components (base frame)
1.1.1 At-a-glance features of the DS8900F
The DS8900F offers the following features (more details can be found in subsequent chapters of this book):
•The DS8900F is available with various POWER9 processor options, from a DS8980F with a 22-core per central processor complex (CPC) option to the DS8950F with a 20-core per CPC option to the DS8910F with an 8-core per CPC option. Each system covers a wide range of performance and cost options.
•The IBM POWER® simultaneous multithreading (SMT) technology allows analytic data processing clients to achieve over 2 million I/O operations per second (IOPS) in database for open environments (70% read and 30% write, 4 KB I/Os, and 50% read cache hit at record-low latencies).
•Memory configurations of 192 GB - 4.3 TB are available.
•The DS8900F uses HPFE Gen2 storage enclosures. One HPFE pair can hold 16, 32, or 48 flash drives, with capacities of 800 GB, 1.6 TB, 3.2 TB, 1.92 TB, 3.84 TB, 7.68 TB, and 15.36 TB. The latter four are more cost-efficient versions of high-capacity flash.
•IBM Z Synergy Services (zsS) include IBM z/OS licensed features that are supported on the storage system. The licensed features include Fibre Channel connection
(IBM FICON®) attachment for IBM Z, parallel access volume (PAV), HyperPAV, SuperPAV, High-Performance FICON for IBM Z (zHPF), zHyperLink, Transparent Cloud Tiering (TCT), z/OS Distributed Data Backup (zDDB), and an option for thin-provisioned Count Key Data (CKD) volumes.
•The DS8900F offers enhanced connectivity with 4-port 32 Gbps and 16 Gbps Fibre Channel Protocol (FCP) / FICON host adapters. Each host adapter port can be configured independently for either FCP or FICON protocol.
•zHPF is an enhancement to the FICON architecture to offload I/O management processing from the IBM Z channel subsystem to the DS8000 host adapters and CPCs. zHPF is an optional feature of IBM Z and DS8000. Recent enhancements to zHPF include Extended Distance Facility zHPF List Pre-fetch support for IBM Db2® and utility operations, and zHPF support for sequential access methods. All Db2 I/O is now zHPF-capable and supports the Db2 castout accelerator function that allows the DS8000 to treat a castout as a single chain of I/Os. zHPF improvements like Extended Distance II further accelerate support for heavy write I/Os over an extended distance of up to 100 km (62 miles).
•DS8900F host adapters include support for IBM Fibre Channel Endpoint Security as part of the cybersecurity solutions that are offered by IBM. Traffic to the host can be authenticated and encrypted. To take advantage of this capability, the platforms that are attached to the DS8900F must also have support for IBM Fibre Channel Endpoint Security. IBM z15 is an example. IBM Fibre Channel Endpoint Security is described in IBM Fibre Channel Endpoint Security for IBM DS8900F and IBM Z, SG24-8455.
•TCT enables direct data movement from the DS8000 CPCs into IBM Cloud® Object Storage, which helps reduce backup workload on IBM Z. Data can be encrypted before leaving the DS8900. The data remains encrypted in cloud storage and is decrypted after it is transferred back to the storage system.
TCT offers seamless hybrid multicloud integration and can be used to move DS8000 data to an IBM TS7700 Virtualization Engine that is configured as an object store. In this case, the data can be encrypted while it is being transferred, and can be compressed. While configured as object storage, the TS7700 also can still be used with traditional FICON logical tape volumes. For more information, see IBM DS8000 Transparent Cloud Tiering (DS8000 Release 9.2), SG24-8381.
•DS8900F includes PCIe-based I/O enclosures. Each I/O enclosure connects to both internal servers over a pair of x8 PCIe Gen3 cables, each of which provides 8 GBps connectivity. Each I/O enclosure also provides two zHyperLink ports.
Each pair of I/O enclosures supports up to two pairs of HPFEs. The storage virtualization that is offered by the DS8900F allows organizations to allocate system resources more effectively and better control application quality of service (QoS).
•The Adaptive Multi-stream Prefetching (AMP) caching algorithm can dramatically improve sequential performance by reducing times for backup, processing for business intelligence, and streaming media. Sequential Adaptive Replacement Cache (SARC) is a caching algorithm that allows you to run different workloads, such as sequential and random workloads, without negatively affecting each other. For example, a sequential workload does not fill up the cache and does not affect cache hits for random workloads. Intelligent Write Caching (IWC) improves the Cache Algorithm for random writes.
•DS8900F provides a graphical management interface to configure, query status information, and monitor the performance of the DS8000. The DS Storage Management GUI (DS GUI) also provides link buttons for accessing the Service Management GUI and Copy Services Manager (CSM) GUI. Additionally, many service functions that once required using the Service Management GUI can now be accessed from the DS GUI.
•IBM Easy Tier® is a no-charge feature that enables automatic dynamic data relocation capabilities. Easy Tier optimizes the usage of the flash storage. No manual tuning is required. The auto-balancing algorithms also provide benefits when homogeneous storage pools are used to eliminate hot spots on arrays of drives.
•Large volume support that allows a DS8000 to support logical unit number (LUN) sizes up to 16 TB. This configuration simplifies storage management tasks. In a z/OS environment, extended address volumes (EAVs) with sizes up to 1 TB are supported.
•The DS8900F has an Active Volume Protection feature that prevents the deletion of volumes that are still in use.
•The American National Standards Institute (ANSI) T10 Data Integrity Field (DIF) Protection Information standard is supported by DS8000 for Small Computer System Interface (SCSI) end-to-end data protection on Fixed-Block (FB) architecture LUN volumes for operating systems (OSs) that can use DIF.
•Dynamic Volume Expansion (DVE) simplifies management by enabling easier, online volume expansion to support application data growth, and to support data center migration and consolidation to larger volumes to ease addressing constraints.
•Thin provisioning allows the creation of over-provisioned devices for more efficient usage of the storage capacity for open systems. For more information, see IBM DS8880 Thin Provisioning (Updated for Release 8.5), REDP-5343.
•Quick Initialization provides fast volume initialization for open system LUNs and CKD volumes. Quick Initialization provides access to volumes the moment that the command completes.
•Full Disk Encryption (FDE), also known as data at rest encryption, can protect business-sensitive data by providing drive-based hardware encryption that is combined with external key management software (IBM Security® Guardium® Key Lifecycle Manager, Gemalto SafeNet KeySecure, or Thales Vormetric Data Security Manager). FDE is available for flash drive types. Because encryption is performed by the drive, it is transparent to host systems, and it can be used in any environment, including z/OS.
•DS8900F R9.2 and later supports a Local Key Management option for FDE on the storage system.
•DS8900F enables clients to comply with SP-800-131a, which is a US National Institute of Standards and Technology (NIST) directive that provides guidance for protecting sensitive data by using cryptographic algorithms that have key strengths of 112 bits. NIST SP-800-131a mode is configured by default in DS8900F.
•Payment Card Industry Data Security Standard (PCI DSS) is a feature of the encryption key management process that helps address payment industry requirements.
•The DS8000 series is certified as meeting the requirements of the IPv6 Ready Logo program. This certification indicates its implementation of IPv6 mandatory core protocols and the ability to interoperate with other IPv6 implementations. The DS8000 can be configured in native IPv6 environments. The logo program provides conformance and interoperability test specifications that are based on open standards to support IPv6 deployment globally. Furthermore, US NIST tested IPv6 with the DS8000 and granted it support for the USGv6 profile and testing program.
•Lightweight Directory Access Protocol (LDAP) authentication support can simplify user management by allowing the DS8000 to rely on a centralized LDAP directory rather than a local user repository. DS8900F supports native LDAP configuration, and it does not require the preinstalled IBM CSM proxy. The CSM proxy can still be used if you want or to facilitate LDAP for older DS8000 systems. For more information, see LDAP Authentication for IBM DS8000 Systems: Updated for DS8000 Release 9.1, REDP-5460.
•For data protection and availability needs, the DS8900F provides a rich set of point-in-time copy (PTC), Remote Mirror, and Remote Copy functions. These functions provide storage mirroring and copying over large distances for disaster recovery (DR) or HA purposes.
•Safeguarded Copy delivers the ability to create and retain hundreds of PTCs to protect against logical data corruption or malicious destruction. Those copies can be used to verify customer data, analyze the nature of the corruption, and restore critical customer data. Dynamic expansion of Safeguarded volume capacity is now possible. For more information, see IBM DS8000 Safeguarded Copy (Updated for DS8000 R9.2), REDP-5506.
•Support for IBM i variable LUN adds flexibility for volume sizes. You can increase capacity usage for IBM i environments. IBM i volumes can be dynamically expanded.
•The DS8900F features Smart Rebuild, a function that is designed to reduce the possibility of secondary failures and data loss in RAID arrays. When a predicted failure occurs on one member of a RAID 5 or RAID 6 array, the affected drive is cloned to a spare drive, allowing it to participate in its own rebuild. The cloning process reduces the duration of the rebuild time. The process falls back to traditional rebuild when Smart Rebuild cannot complete.
•The DS8900F offers enhanced IBM Z Synergy items:
 – Forward Error Correction (FEC)
FEC is a protocol that detects and corrects bit errors that are generated during data transmission. Both the IBM z13® (and later) and the DS8000 extend the usage of FEC to complete end-to-end coverage for 16 Gbps and faster links and preserve data integrity with more redundancy.
 – Fibre Channel (FC) Read Diagnostic Parameters (RDP)
FC RDP improves the end-to-end link fault isolation for 16 Gbps and faster links on current IBM Z and DS8000 systems. RDP data provides the optical signal strength, error counters, and other critical information that is crucial to determine the quality of the link.
 – FICON Dynamic Routing (FIDR)
FIDR is an IBM Z and DS8000 feature that supports the use of dynamic routing policies in the fabric switch to balance load across inter-switch links (ISLs) on a per I/O basis.
 – Fabric I/O Priority
Fabric I/O Priority is an end-to-end synergy feature among the z/OS Workload Manager (WLM), Brocade storage area network (SAN) Fabric, and the storage system to manage QoS on a single I/O level.
 – IBM Fibre Channel Endpoint Security
As one of the first hosts, and following a pervasive encryption philosophy, IBM z15™ supports SAN authentication and FC encryption of FICON traffic to the DS8900F.
 – IBM zHyperLink
zHyperLink is a short distance link technology that dramatically reduces latency by interconnecting the IBM Z CPCs directly to the I/O bays of the DS8900.
All DS8900F models can support zHyperLink. The current zHyperLink release supports read/write I/O. zHyperLink is intended to complement FICON technology to accelerate I/O requests that are typically used for transaction processing.
For more information, see Getting Started with IBM zHyperLink for z/OS, REDP-5493.
 
Note: For more information about the IBM Z synergy features, see IBM DS8900F and IBM Z Synergy DS8900F: Release 9.3 and z/OS 2.5, REDP-5186.
1.2 DS8900F controller options and frames
The DS8900F family consists of a series of distinct models that are all flash storage:
•DS8980F machine type 5341 model 998 and DS8980F model E96
•DS8950F machine type 5341 model 996 and DS8950F model E96
•DS8910F machine type 5341 model 994 as racked in a single frame
•DS8910F machine type 5341 model 993 as the rack-mounted model
The DS8900F includes the following features:
•IBM POWER9 processor-based server technology
The DS8900F features the IBM POWER9 processor-based server technology for high performance. Compared to the IBM POWER8® processors that were used in the DS8880 servers, IBM POWER9 processor-based servers excel with a smaller front-end latency and up to 60% more IOPS in transaction-processing workload environments. The number of transistors per core has almost doubled between these two POWER generations.
•Nondisruptive upgrade path
A nondisruptive upgrade path for the DS8900F allows processor, cache, host adapters, storage upgrades, and expansion frame upgrades to be installed concurrently without disrupting applications.
•The air-flow system allows optimal horizontal cool down of the storage system. The DS8900F is designed for hot- and cold-aisle data center design, drawing air for cooling from the front of the system and exhausting hot air at the rear. For more information, see 2.6, “Power and cooling” on page 58.
•The DS8900F offers three system class options:
 – DS8980F Mod 998 - Analytic Class, which is based on the IBM Power S924 architecture.
 – DS8950F Mod 996 - Agility Class, which is based on the IBM Power S924 architecture.
 – DS8810F Mod 994 and Mod 993 - Flexibility Class, which is based on IBM Power S922 architecture.
For more information about the specifics for each model, see 2.3.1, “IBM POWER9 processor-based CPCs” on page 42.
•HPFE
HPFE Gen2 is an enclosure that can support both high-performance flash drives with capacities of 800 GB - 3.2 TB and high-capacity flash drives of 1.92 TB, 3.84 TB, 7.68 TB, or 15.36 TB (2.5-inch form-factor) in a 2U rack space. The enclosures must be installed in pairs.
For more information, see IBM DS8000 High-Performance Flash Enclosure Gen2 (DS8000 R9.0), REDP-5422.
•Rack-mountable and racked options
The DS8900F comes in racks with a footprint that is reduced compared its DS8880 predecessor models. To use the DS8900F in a customer-provided rack, use the DS8910F model 993.
1.3 DS8900F architecture and functions overview
The DS8900F offers continuity with the fundamental architecture of its predecessors: the DS8880 and DS8870 models. This architecture ensures that the DS8900F can use a stable and proven operating environment that offers optimal availability. The hardware is optimized to provide higher performance, connectivity, and reliability.
1.3.1 Overall architecture and components
For more information about the available configurations for the DS8900F, see Chapter 2, “IBM DS8900F hardware components and architecture” on page 25.
 
Note: Some technical aspects are specific to the DS8910F rack-mounted model. For a full overview of the architectural aspects of the DS8910F rack-mounted model, see IBM DS8910F Model 993 Rack-Mounted Storage System Release 9.1, REDP-5566.
IBM POWER9 processor technology
The POWER9 processor is manufactured by using the IBM 14 nm Silicon-On-Insulator (SOI) technology. Each chip is 693 mm2 and contains 8 billion transistors. The DS8900F uses the PCIe I/O controllers and an interconnection system that connects all components within the chip. POWER9 processor advancements in multi-core and multithreading are remarkable. These multithreading capabilities improve the I/O throughput of the DS8900F storage servers.
The DS8900F family offers several configurations of CPCs:
•The DS8980F model has two 22-core CPCs (each CPC has two 11-core processors) with a total of 4.3 TB of system memory.
•The DS8950F offers two configurations:
 – 10-core per CPC with a total of 512 GB of total system memory
 – 20-core per CPC with a total of 1, 2, or 3.4 TB of total system memory
•The DS8910F models have an 8-core processor configuration per CPC with 192 GB or 512 GB of system memory.
A CPC is also referred to as a storage server. For more information, see Chapter 3, “IBM DS8900F reliability, availability, and serviceability” on page 71.
Internal PCIe-based fabric
The DS8900F fabric includes the following specifications:
•DS8900F POWER9 processor-based servers are based on the current PCIe Gen4 architecture to provide up to 16-lane (x16) high-speed connectivity to internal adapters.
•PCIe adapters provide point-to-point connectivity to the I/O enclosures. The I/O enclosures provide connectivity between the I/O adapters and the POWER9 processor complexes.
•The I/O enclosures provide PCIe Gen3 connectivity to all installed host and device adapters. Each I/O enclosure features six PCIe x8 adapter slots and two extra PCIe x8 connectors for attachment to zHyperLink optical transceivers.
HPFE Gen2
The HPFE flash RAID adapters are installed in pairs and split across an I/O enclosure pair. They occupy the third and sixth PCIe slots according to the adapter pair plugging order.
HPFE drive enclosures are also installed in pairs, and connected to the corresponding flash RAID adapter pair over eight 6 Gbps SAS cables for high bandwidth and redundancy. Each drive enclosure can contain up to twenty-four 2.5-inch SAS flash drives. Flash drives are installed in groups of 16, and split evenly across the two drive enclosures in the pair.
Each flash adapter pair and HPFE pair deliver up to 900 K IOPS reads, 225 K IOPS writes, and up to 14 GBps (read) and 9.5 GBps (write) bandwidth.
Drive options
Flash drives provide up to 100 times the throughput and 10 times lower response time than 15 K revolutions per minute (RPM) hard disk drives (HDDs). They also use less power than traditional HDDs. For more information, see Chapter 5, “IBM DS8900F physical planning and installation” on page 141.
Flash drives are grouped into three tiers, based on performance and capacity. These flash Drives are supported across all DS8900F models.
•2.5-inch flash Tier 0 high-performance flash drives:
 – 800 GB
 – 1.6 TB
 – 3.2 TB
•2.5-inch flash Tier 1 high-capacity flash drives:
 – 3.84 TB
•2.5-inch flash Tier 2 high-capacity flash drives:
 – 1.92 TB
 – 7.68 TB
 – 15.36 TB
All flash drives in the DS8900F are encryption-capable. Enabling encryption is optional, and requires at least two external key servers or the local key management feature.
Easy Tier
Easy Tier enables the DS8000 to automatically balance data placement on disk drives to avoid hot spots on flash arrays. Easy Tier can place data in the storage tier that best suits the access frequency of the data. Highly accessed data can be moved nondisruptively to a higher tier, for example, to 1.6 TB Flash Tier 0 drives. Cold data or data that is primarily accessed sequentially is moved to a lower tier of high-capacity drives. Easy Tier includes more components:
•Easy Tier Application is an application-aware storage utility to help deploy storage more efficiently by enabling applications and middleware to direct more optimal placement of the data by communicating important information about current workload activity and application performance requirements. It is possible for Db2 applications in z/OS environments to give hints of data placement to Easy Tier at the data set level.
•Easy Tier Heat Map Transfer (HMT) can take the data placement algorithm on the Metro Mirror (MM) Global Copy (GC) and Global Mirror (GM) primary site and reapply it to the MM, GC, or GM secondary site when failover occurs by using the Easy Tier Heat Map Transfer Utility (HMTU). With this capability, the DS8000 systems can maintain application-level performance. The Easy Tier HMT functions support Metro/Global Mirror (MGM) to transfer a heat map automatically to a tertiary site.
 
Note: Easy Tier Server was removed from marketing support. It was replaced by the Flash Cache option of IBM AIX® 7.2.
The following Easy Tier controls are also available:
•For Easy Tier to function effectively, it needs some free extents in each extent pool to be able to move around extents. Rather than having the storage administrator monitor capacity, the system can reserve some space for Easy Tier extent movements. You can enable space reservation for Easy Tier by running chsi -etsrmode enable (enabled by default).
•Another control is the allocation policy for new volumes. You can control the allocation policy according to your needs by running the chsi command with -ettierorder highutil or highperf. The data allocation order for all flash systems can be changed between the following settings:
 – High-Utilization: Flash Tier 1 → Flash Tier 2 → Flash Tier 0
 – High-Performance: Flash Tier 0 → Flash Tier 1 → Flash Tier 2 (the default)
For more information about Easy Tier features, see the following resources:
•IBM DS8000 Easy Tier (Updated for DS8000 R9.0), REDP-4667
•DS8870 Easy Tier Application, REDP-5014
•IBM DS8870 Easy Tier Heat Map Transfer, REDP-5015
Host adapters
The DS8900F offers 32 Gbps and 16 Gbps host adapters. Both types have four ports each, and each port can be independently configured for either FCP or FICON:
•The 32 Gbps adapter has four ports. Each port independently auto-negotiates to an 8, 16, or 32 Gbps link speed.
•The 16 Gbps adapter has four ports. Each port independently auto-negotiates to a 4, 8, or 16 Gbps link speed.
Storage Hardware Management Console for the DS8000
HMCs are the focal point for notification, management, and maintenance activities. HMCs proactively monitor the state of your system and notify you and IBM when service is required. HMCs can also be used for management of copy services by using the preinstalled IBM Copy Services Manager (CSM) software.
Every DS8900F includes two HMCs for redundancy, which are installed in the management enclosure in the base rack. DS8900F HMCs support IPv4 and IPv6 standards. For more information, see Chapter 6, “IBM DS8900F Management Console planning and setup” on page 167.
Isolated key server
The DS8900F includes FDE flash drives. To configure a DS8900F to use data at rest encryption, at least two key servers are required. An isolated key server with dedicated hardware and non-encrypted storage resources is required. It can be ordered from IBM. For more information, see 5.3.6, “Key manager servers for encryption” on page 161.
You can also encrypt data before it is transmitted to the cloud when using the TCT feature. For more information, see the IBM DS8000 Encryption for Data at Rest, Transparent Cloud Tiering, and Endpoint Security (DS8000 Release 9.2), REDP-4500.
Local Key Manager
Release 9.2 and 9.3 offer Local Key Management for encryption. Local Key Management provides a DS8000 encryption and key management solution to minimize the risk of exposure and reduce operational costs. Local key encryption offers good security for data on disk, even when a drive is physically removed with the intent of stealing the data. For more information about local key manager, see IBM DS8000 Encryption for Data at Rest, Transparent Cloud Tiering, and Endpoint Security (DS8000 Release 9.2), REDP-4500.
1.3.2 Storage capacity
The physical storage capacity for the DS8900F is installed in fixed increments that are called drive sets or flash drive sets. A drive set contains 16 flash drives, all of which have the same capacity and the same drive class. Both high-performance and high-capacity flash drives are available in sets of 16.
The available drive options provide industry-class capacity and performance to address a wide range of business requirements. The DS8000 storage arrays can be configured as
RAID 6, RAID 10, or RAID 5, depending on the drive type.
RAID 6 is now the default and preferred setting for the DS8900F. RAID 5 can be configured for drives of less than 1 TB, but this configuration is not preferred and requires a risk acceptance, and a field Request for Price Quotation (RPQ) for enterprise hard disk drive (HDD) drives. Flash Tier 0 drive sizes larger than 1 TB can be configured by using RAID 5, but require an RPQ and an internal control switch to be enabled. RAID 10 continues to be an option for all-flash drives.
1.3.3 Supported environments
The DS8000 offers connectivity support across a broad range of server environments, including IBM Power, IBM Z, servers from HPE and Oracle, AMD-based, and Intel-based x64 servers.
The DS8000 supports over 60 platforms. For the list of supported platforms, see the IBM System Storage Interoperation Center (SSIC) for DS8000.
This rich support of heterogeneous environments and attachments, with the flexibility to partition easily the DS8000 storage capacity among the attached environments, can help support storage consolidation requirements and dynamic environments.
1.3.4 Configuration flexibility
The DS8000 series uses virtualization techniques to separate the logical view of hosts onto LUNs from the underlying physical layer, providing high configuration flexibility. For more information about virtualization, see Chapter 4, “Virtualization concepts” on page 107.
Dynamic Volume Expansion
DVE increases the capacity of open systems, IBM i, and IBM Z volumes while the volumes remain connected to a host system. This capability simplifies data growth by providing volume expansion without taking volumes offline. Certain OSs do not support a change in volume size. Therefore, a host action is required to detect the change after the volume capacity is increased.
Large LUN and large Count Key Data volume support
You can configure LUNs and volumes to span arrays, allowing for larger LUN sizes of up to
16 TB in open systems.
 
Tip: Copy Services (CS) are currently supported for LUN sizes of up to 4 TB.
The maximum CKD volume size is 1,182,006 cylinders (1 TB), which can greatly reduce the number of volumes that are managed. This large CKD volume type is called a 3390 Model A. It is referred to as an Extended Address Volume (EAV).
T10 Data Integrity Field support
The DS8900F supports the T10 DIF standard for FB volumes that are accessed by the FCP channel of Linux on IBM Z and IBM AIX on IBM Power. You can define LUNs with an option to instruct the DS8000 to use the CRC-16 T10 DIF algorithm to store the data. You can also create T10 DIF-capable LUNs. The support for IBM i variable LUNs adds flexibility for volume sizes and can increase capacity usage for IBM i environments.
VMware vStorage API for Array Integration support
The VMware vStorage API for Array Integration (VAAI) feature offloads specific storage operations to the storage system for greatly improved performance and efficiency. With VAAI, VMware vSphere can perform key operations faster and use less CPU, memory, and storage bandwidth.
The DS8000 supports the following VAAI primitives:
•Atomic Test and Set (ATS), which is also known as Compare and Write for hardware-assisted locking.
•Clone Blocks (Extended Copy or XCOPY) for hardware-assisted move or cloning. For XCOPY, the DS8000 CS license is required. Also, XCOPY is not supported by extent space efficient (ESE) volumes and volumes larger than 4 TB; the target of an XCOPY cannot be an MM or GM primary volume.
•Write Same (or Block Zero) is used to initialize new volumes and is supported by all volumes.
•Space Release (UNMAP) is supported only with ESE volumes that use small extents
(16 mebibytes (MiB)).
IBM DS8000 Storage Replication Adapter
IBM DS8000 Storage Replication Adapter (SRA) is a software add-on that integrates with the VMware vCenter Site Recovery Manager (SRM) solution and enables SRM to perform failovers together with DS8000 storage systems. The DS8000 SRA extends SRM capabilities and allows it to employ DS8000 replication and mirroring as part of the SRM comprehensive disaster recovery planning (DRP) solution.
OpenStack
The DS8000 supports the OpenStack cloud management software for business-critical private, hybrid, and public cloud deployments. The DS8900F supports features in the OpenStack environment, such as volume replication and volume retype. The Cinder driver for DS8000 is now open source in the OpenStack community. The /etc/cinder.conf file can be directly edited for the DS8000 back-end information.
For more information about the DS8000 and OpenStack, see Using IBM DS8000 in an OpenStack Environment, REDP-5220.
Red Hat OpenShift
Red Hat OpenShift is an open source container application platform that is based on the Kubernetes container orchestrator for enterprise application development and deployment. IBM Red Hat OpenShift Container Platform (OCP) provides developers and IT organizations with a hybrid cloud application platform for deploying new and existing applications on secure, scalable resources with minimal configuration and management overhead. OCP supports various programming languages and frameworks, such as Java, JavaScript, Python, Ruby, and PHP.
DS8900F supports the Container Storage Interface (CSI) specification. IBM released an open source CSI driver for IBM storage that allows dynamic provisioning of storage volumes for containers on Kubernetes and IBM Red Hat OpenShift Container Platform (OCP).
The CSI driver for IBM block storage systems enables container orchestrators such as Kubernetes to manage the lifecycle of persistent storage. This CSI is the official operator to deploy and manage the IBM block storage CSI driver.
For more information about CSI, see IBM Documentation.
RESTful application programming interface
With the DS8000 support of RESTful application programming interface (API) services, DS8900F clients or cloud administrators can design and implement the DS8000 management applications by using the Representational State Transfer (REST) software architecture.
For more information about the RESTful API, see Exploring the DS8870 RESTful API Implementation, REDP-5187.
Flexible LUN-to-LSS association
With no predefined association of arrays to logical subsystems (LSSs) on the DS8000 series, users can put LUNs or CKD volumes into LSSs and better use the 256 address range, particularly for IBM Z.
Simplified LUN masking
In the new GUI, LUNs are directly mapped to the host, and the user cannot define volume groups. Volume groups still exist on the DS8900F systems, but they are not visible from the GUI because they are created in the background during the assignment of a volume to the host.
Thin-provisioning features
Volumes in the DS8900F can be provisioned as full or thin. When clients plan capacity, they must consider the number of volumes in the extent pool (or overall storage system) and the degree of over-allocation that is planned for.
These volumes feature enabled over-provisioning capabilities that provide more efficient usage of the storage capacity and reduced storage management requirements. For more information, see Chapter 4, “Virtualization concepts” on page 107 and IBM DS8880 Thin Provisioning (Updated for Release 8.5), REDP-5343.
Maximum values of logical definitions
The DS8000 features the following maximum values for the major logical definitions:
•Up to 255 LSSs
•Up to 65,280 logical devices
•Up to 16 TiB LUNs
•Up to 1,182,006 cylinder (1 TB) CKD volumes
•Up to 130,560 FICON logical paths (512 logical paths for each control unit image) on the DS8000
•Up to 1,280 logical paths for each FC port
•Up to 8,192 process logins (509 for each SCSI-FCP port)
1.3.5 Copy Services functions
For IT environments that cannot afford to stop their systems for backups, the DS8000 provides IBM FlashCopy®. This fast replication technique can provide a PTC of the data in a few seconds or even less. A recent addition is the ability to perform Cascaded FlashCopy backups, where the target of a FlashCopy relationship can also be the source of another FlashCopy relationship. For more information, see DS8000 Cascading FlashCopy Design and Scenarios, REDP-5463.
To meet the challenges of cybersecurity, the Safeguarded Copy function, based on the FlashCopy technology, can create and retain hundreds of PTCs for protection against logical data corruption. Release 9.2 added the capability to restore a recovered Safeguarded Copy to a production copy of the data. For more information, see IBM DS8000 Safeguarded Copy (Updated for DS8000 R9.2), REDP-5506.
 
For data protection and availability needs, the DS8900F provides MM, GM, GC, MGM, and z/OS Global Mirror (zGM), which are Remote Mirror and Remote Copy functions. These functions are also available and are fully interoperable with previous models of the DS8000 family. These functions provide storage mirroring and copying over large distances for DR or availability purposes.
CS scope limiting is the ability to specify policy-based limitations on CS requests.
For more information about CS, see IBM DS8000 Copy Services: Updated for IBM DS8000 Release 9.1, SG24-8367.
1.3.6 Service and setup
Beginning with Release 9.3, DS8000 systems have a single machine type that is called 5341. The former warranty and service options are now offered as part of Expert Care. Options range from a 1-year base warranty to 5-year Expert Care Advanced or Premium, which may be chosen when you order a new system or upgrade as requirements change.
Figure 1-4 describes the available Expert Care options.
Figure 1-4 Expert Care: 1-year and 5-year options
Physical installation of the DS8000 is performed by IBM by using the installation procedure for this system. The client’s responsibility is the installation planning, retrieval, and installation of feature activation codes, logical configuration, and execution.
The storage system HMC is the focal point for maintenance and service operations. Two HMCs are inside the DS8900F management enclosure, and they continually monitor the state of the system. HMCs notify IBM, and they can be configured to notify you when service is required.
The HMC is also the interface for Call Home and remote support, which can be configured to meet client requirements. It is possible to allow one or more of the following configurations:
•Call Home on error (machine-detected)
•Remote connection for a few days (client-initiated)
•Remote problem log collection (service-initiated)
The remote connection between the HMC and IBM Support is performed by using the Assist On-site (AOS) feature. AOS offers more options, such as Transport Layer Security (TLS), and enhanced audit logging. For more information, see IBM Assist On-site for Storage Overview, REDP-4889.
IBM Remote Support Center (RSC) is also an available option for providing IBM Support remote support access to systems.
For Release 9.3 systems, IBM provides three options for microcode updates:
•Customer Code Load
•Remote Code Load (RCL)
•Onsite SSR Code Load
For customers who choose the base warranty service or Expert Care Advanced, Customer Code Load is the default method for performing concurrent microcode updates:
•Microcode bundles are downloaded and activated by the customer by using the standard DS Storage Manager GUI.
•The download defaults to the current recommended bundle, or an alternative compatible bundle may be chosen.
•Health checks are run before the download, and again before activation to ensure that the system is in good health.
•If a problem is encountered anywhere in the process, a ticket is opened automatically with IBM Support, and the ticket number is provided in the GUI for reference.
•After the problem is corrected, the code load can be restarted, and automatically resumes after the last successful step.
Customers who want to have an IBM Systems Service Representative (IBM SSR) perform Onsite Code Load may purchase Feature Code #AHY2 with Expert Care Premium, or Feature Code #AHY3 with Expert Care Advanced.
1.3.7 IBM Certified Secure Data Overwrite
IBM Certified Secure Data Overwrite (SDO) is a process that provides a secure overwrite of all data storage in a DS8900F storage system. Before you perform a SDO, you must remove all the logical configuration and any encryption groups that may be configured. The process is then initiated by the IBM SSR. For more information, see 3.7.3, “IBM Certified Secure Data Overwrite” on page 101.
1.4 Performance features
The DS8900F offers optimally balanced performance. The DS8000 incorporates many performance enhancements, such as a dual multi-core IBM POWER9 processor complex implementation, fast 32 Gbps and 16 Gbps FCP / FICON host adapters, HPFE dedicated flash architecture in second generation, classical flash drives, and high-bandwidth, fault-tolerant point-to-point PCIe internal connections.
With all of these components, the DS8900F is positioned at the top of the high-performance category.
1.4.1 4.3 TB cache
DS8900F supports up to a 4.3 TB cache, which can improve the cache hit ratio and result in the reduction of the response time of the same workload demand. For a cache size sensitive workload with an improved cache hit ratio, the larger cache size provides higher throughput with lower response time.
1.4.2 32 Gbps and 16 Gbps host adapters
The DS8900F supports 32 Gbps and 16 Gbps host adapters. This connectivity reduces latency and provides faster single stream and per-port throughput. These adapters can work up to two-speed classes below their own nominal class, but they do not support Fibre Channel Arbitrated Loop (FC-AL) connections. The Lights on Fastload feature avoids path disturbance during code loads.
The 16 Gbps host bus adapter (HBA) supports IBM Fibre Channel Endpoint Security authentication. The 32 Gbps HBA supports both IBM Fibre Channel Endpoint Security authentication and line-rate encryption.
1.4.3 Sophisticated caching algorithms
IBM Research® conducts extensive investigations into improved algorithms for cache management and overall system performance improvements. To implement sophisticated caching algorithms, it is essential to include powerful processors for the cache management. With a 4 KB cache segment size and up to 4.3 TB overall cache sizes, the tables to maintain the cache segments become large.
Sequential Prefetching in Adaptive Replacement Cache
One of the performance features of the DS8000 is its self-learning cache algorithm, which optimizes cache efficiency and enhances cache hit ratios. This algorithm, which is used in the DS8000 series, is called Sequential Prefetching in Adaptive Replacement Cache (SARC).
SARC provides the following abilities:
•Sophisticated algorithms to determine the data to store in cache that is based on recent access and the frequency needs of the hosts.
•Prefetching, which anticipates data before a host request and loads it into cache.
•Self-learning algorithms to adaptively and dynamically learn the data to store in cache that is based on the frequency needs of the hosts.
Adaptive Multi-stream Prefetching
AMP is a breakthrough caching technology that improves performance for common sequential and batch processing workloads on the DS8000. AMP optimizes cache efficiency by incorporating an autonomic, workload-responsive, and self-optimizing prefetching technology.
Intelligent Write Caching
IWC improves performance through better write-cache management and destaging the order of writes. IWC can also double the throughput for random write workloads. Specifically, database workloads benefit from this new IWC cache algorithm.
SARC, AMP, and IWC play complementary roles. Although SARC carefully divides the cache between the RANDOM and the SEQ lists to maximize the overall hit ratio, AMP is managing the contents of the SEQ list to maximize the throughput that is obtained for the sequential workloads. IWC manages the write cache and decides the order and rate to destage to disk.
1.4.4 Flash storage
Today’s installations are mostly hybrid multitier installations, mixing more cost-economical large-capacity drives with high-performance flash together in one system. Given the capacity increase of the flash modules, the price drop of capacity-optimized flash compared to high-RPM HDDs, and given their savings on energy consumption and space, most clients decide on an all-flash array storage strategy.
The DS8900F flash storage can be tiered, with three tiers of flash storage that are available. Then, you can use Easy Tier to optimize the storage. The DS8900F offers automated algorithms that optimize the tiering and place hot areas onto higher-tiered flash arrays.
To improve data transfer rate (IOPS) and response time, the DS8900F supports flash drives and high-performance flash drives, which are based on NAND technology. With the flash drives and the specific architecture that is used in the HPFEs, much higher IOPS densities (IOPS per GB) are possible than with ordinary solid-state drives (SSDs).
Flash drives sharply improve I/O transaction-based performance over traditional HDDs in standard drive enclosures.
High-performance flash drives use the flash RAID adapters in the I/O enclosures, and PCIe connections to the processor complexes. The high-performance flash drive types are high-IOPS class enterprise storage devices that are targeted at flash Tier 0, for I/O-intensive workload applications that need high-level, fast-access storage. The high-capacity flash drive types for flash Tiers 1 and 2 often have an acquisition price point that helps eliminate HDDs when replacing a storage system.
Flash drives offer many potential benefits over HDDs, including better IOPS, lower power consumption, less heat generation, and lower acoustical noise. For more information, see Chapter 5, “IBM DS8900F physical planning and installation” on page 141.
1.4.5 Performance for IBM Z
The DS8000 series supports the following IBM performance enhancements for IBM Z environments:
•PAVs enable a single IBM Z server to simultaneously process multiple I/O operations to the same logical volume, which can reduce device queue delays. This reduction is achieved by defining multiple addresses for each volume. With Dynamic PAV, the assignment of addresses to volumes can be automatically managed to help the workload meet its performance objectives and reduce overall queuing.
•HyperPAV enables applications to achieve equal or better performance than with PAV alone while using fewer unit control blocks (UCBs) and eliminating the latency in targeting an alias to a base. With HyperPAV, the system can react immediately to changing I/O workloads.
•SuperPAV allows z/OS to use an alias address from another logical control unit (LCU).
•Multiple Allegiance expands the simultaneous logical volume access capability across multiple IBM Z servers. This function, with PAV, enables the DS8000 to process more I/Os in parallel, which improves performance and enables greater use of large volumes.
•I/O priority queuing allows the DS8000 series to use I/O priority information that is provided by the z/OS WLM to manage the processing sequence of I/O operations at the adapter level.
•zHPF enables multiple channel commands to be sent to the control unit as a single entity. The channel forwards a chain of commands, and it does not need to track each individual channel command word (CCW). This configuration improves FICON throughput on the DS8000 I/O ports. The DS8000 systems also support the new zHPF I/O commands for multi-track I/O operations, Db2 list-prefetch, sequential access methods, and Db2 castout acceleration.
•zHyperWrite is another enhancement for Db2 clients. In a z/OS MM environment, it enables Db2 log updates to be written to the primary and secondary volumes in parallel. This configuration reduces the latency for log writes, and so improving transactional response times and log throughput. The MM primary volume needs to be enabled with IBM HyperSwap® by either IBM Geographically Dispersed Parallel Sysplex
(IBM GDPS) or IBM Copy Services Manager.
•zHyperLink provides a connectivity method that dramatically reduces latency by interconnecting the IBM Z system directly to the I/O bay of the DS8900. Response times lower than 20 µs are possible for qualifying I/Os. zHyperLink supports read/write I/O.
•zHyperLink writes to the GM primary extends the benefits of zHyperLink to GM environments. With 32 Gbps FICON connectivity, where a zHPF write operation to a GM takes about 100 µs, that same write operation takes only 27 µs over zHyperLink.
•Release 9.2 added the ability to allow consistent read from MM secondary volumes over zHyperLink or zHPF in a HyperSwap configuration.
•With the TCT function, the DS8000 CPCs can be scheduled to directly move data over to Cloud Object Storage or to tape without data going through the host, which can sharply reduce IBM Z MIPS usage during the backup periods and reduce expenses while simplifying data archiving operations. The TCT function now enables multi-cloud support where up to eight cloud targets can be defined for each DS8900F. For more information, see IBM DS8000 Transparent Cloud Tiering (DS8000 Release 9.2), SG24-8381.
For more information about performance on IBM Z, see IBM DS8900F and IBM Z Synergy DS8900F: Release 9.3 and z/OS 2.5, REDP-5186.
1.4.6 Performance enhancements for IBM Power servers
Many IBM Power users can benefit from the following DS8000 features:
•End-to-end I/O priorities
•Cooperative caching
•Long busy wait host tolerance
•Automatic port queues
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
18.226.104.250