Introducing the IBM TS7700
The IBM TS7700, which was introduced in 2006, is now in its sixth generation of IBM Tape Virtualization products for mainframes. It replaces the highly successful IBM TotalStorage Virtual Tape Server (VTS).
This chapter includes the following topics:
1.1 Overview
With cloud infrastructures on the rise and data volumes that are expanding exponentially, organizations need a cost effective way to manage both primary and backup data that is active, inactive, or even archived. Long-term retention of data is a business priority, as is continuous availability from anywhere at any time, and the storage solution must also fit within today’s budget constraints. Storing infrequently accessed data on costly disk storage simply does not make sense. At the same time, physical tape libraries can require long access times, making the use cost-prohibitive in transactional storage infrastructures. That’s where virtualized tape storage comes in.
This publication explains what is introduced with IBM TS7700 release R4.2 and the concepts associated with it. TS7700 R4.2 can be installed on the IBM TS7720, TS7740, and TS7760 models only. The IBM TS7720T and TS7760T (tape attach) partition mimics the behavior of the previous TS7740, but with higher performance and capacity. R4.2 introduces the ability of the TS7760 to store virtual tape volumes to an object store.
|
Note: For more information about the TS7760 Cloud Object Storage solution and how to implement and integrate this solution into your enterprise, see IBM TS7760 R4.2 Cloud Storage Tier Guide, REDP-5514:
|
The sixth-generation TS7700 consists of the following models:
•Current models:
– TS7760C (cloud storage tier)
– TS7760T (tape-attached)
– TS7760 VEC (disk only, upgradeable to TS7760T or TS7760C)
•Previous models:
– TS7740 V07 (tape-attached)
– TS7720T (tape-attached)
– TS7720 VEB (disk only)
The TS7700 is a modular, scalable, and high-performance architecture for mainframe tape virtualization. This architecture is a fully integrated, tiered storage hierarchy of disk, tape, and now cloud object stores. It incorporates extensive self-management capabilities consistent with IBM Information Infrastructure initiatives.
These capabilities can improve performance and capacity. Better performance and capacity help lower the total cost of ownership for tape processing, and help avoid human error. A TS7700 can improve the efficiency of mainframe tape operations by efficiently using disk storage, tape capacity, and tape speed. It can also improve efficiency by providing many tape addresses.
TS7700 provides tape virtualization for the IBM z environment. Tape virtualization can help satisfy the following requirements in a data processing environment:
•Improved reliability and resiliency
•Reduction in the time that is needed for the backup and restore process
•Reduction of services downtime that is caused by physical tape drive and library outages
•Reduction in cost, time, and complexity by moving primary workloads to virtual tape
•More efficient procedures for managing daily backup and restore processing
•Infrastructure simplification through reduction of the number of physical tape libraries, drives, and media
1.2 New capabilities
Building on over 20 years of virtual tape experience, the TS7760 now supports the ability to store virtual tape volumes in an object store. The TS7700 supported off loading to physical tape for over two decades. Off loading to physical tape behind a TS7700 is utilized by hundreds of organizations around the world.
By using the same hierarchical storage techniques, the TS7700 can also off load to object storage. Because object storage is cloud-based and accessible from different regions, the TS7760 Cloud Storage Tier support essentially allows the cloud to be an extension of the grid. As of this writing, the TS7760C supports the ability to off load to IBM Cloud Object Storage and Amazon S3.
TS7700 4.2 delivers the following capabilities:
•8-way Grid Cloud, which consists of any generation of TS7700
•Synchronous and asynchronous replication
•Tight integration with IBM Z and DFSMS policy management
•Optional Transparent Cloud Tiering
•Optional integration with physical tape
•Cumulative 16 Gb FICON throughput up to 4.8 GBps * 8
•IBM Z hosts view up to 496 * 8 equivalent devices
•Grid access to all data independent of where it exists
1.3 Concepts of storage virtualization
A virtual tape subsystem presents emulated tape drives to the host, and stores tape data on emulated tape volumes. These volumes are stored in a disk-based cache rather than physical tape media. The TS7700 emulates the function and operation of IBM 3490 Enhanced Capacity (3490E) tape drives. Depending on the TS7700 model, it uses a Redundant Array of Independent Disks (RAID) technology or Dynamic Disk Pools to store volumes in cache that are written by the host. The disk space that is provided is called a Tape Volume Cache (TVC).
The main components of the IBM TS7700 are shown in Figure 1-1.
Figure 1-1 Main components of the TS7700
Emulated tape drives are also called virtual drives. To the host, virtual IBM 3490E tape drives look the same as physical 3490E tape drives. Emulation is not apparent to the host and applications. The host always writes to and reads from virtual tape drives. It never accesses the physical tape drives (commonly referred to as the back-end tape drives) attached to TS7740, TS7720T, and TS7760T configurations. In fact, it does not need to identify that these tape drives exist. Nor does the host ever directly access object stores in the cloud for the TS7760C or have awareness that volumes are stored there.
Even an application that supports only 3490E tape technology can use the TS7700 without any changes. Therefore, the application benefits from the high capacity and high-performance tape drives in the back end. For TS7720 VEB and TS7760 VEC (disk-only) configurations, no physical tape attachment exists. However, the virtual tape drives work the same for the host.
Because the host exclusively accesses the virtual tape drives, all data must be written to or read from emulated volumes in the disk-based TVC. These emulated tape volumes in the TVC are called virtual volumes.
When the host requests a volume that is still in disk cache, the volume is virtually mounted. No physical mount is required. After the virtual mount is complete, the host can access the data at disk speed. Mounting scratch tapes is also virtual, and does not require a physical mount.
Although you define maximum sizes for your volumes, a virtual volume takes up only the space in cache that the compressed data on the volume requires. For this reason, tape virtualization makes efficient use of disk capacity. In IBM TS7740, IBM TS7720T, and TS7760T (tape-attached) configurations, the virtual volumes are copied from disk to tape. They also need only the amount of tape capacity that is occupied by the data that is stacked end-to-end, which makes efficient use of disk and tape capacity. In the IBM TS7760C (cloud storage tier) configuration, the virtual volumes are copied from disk to object stores in the cloud that is based on the actual number of bytes of data rather than the volume size.
Another benefit of tape virtualization is the large number of drives available to applications. Each IBM TS7700 can support up to a maximum of 496 virtual tape devices. Often, applications contend for tape drives, and jobs must wait because no physical tape drive is available. Tape virtualization efficiently addresses these issues by providing many virtual tape drives. The TS7740, TS7720T, and TS7760T manage the physical tape drives and physical volumes in the tape library. It also controls the movement of data between physical and logical volumes.
In the TS7740, TS7720T, TS7760T, and TS7760C, data that is written from the host into the TVC is scheduled for copying to tape or cloud later. The process of copying data to tape or cloud that exists only in cache is called premigration. When a volume is copied from cache to tape or cloud, the volume is called a logical volume.
A physical volume can contain many logical volumes. The process of putting several logical volumes on one physical tape is called stacking. A physical tape that contains logical volumes is referred to as a stacked volume. This concept does not apply to TS7720 VEB and TS7760 VEC because no physical tape devices are attached to it.
Without a TS7740, TS7720T, and TS7760T, many applications would be unable to fill the high capacity media of modern tape technology, and you might end up with many under-used cartridges. This wastes much space, and requires an excessive number of cartridge slots.
Tape virtualization eliminates any unused volume capacity, and fully uses physical tape capacity when present. Also, you can use tape virtualization to use the full potential of modern tape drive and tape media technology. In addition, it does so without changes to your applications or job control language (JCL).
When space is required in the TVC of a TS7740, TS7720T, TS7760T, and TS7760C for new data, volumes that were copied to tape or cloud are removed from the cache. By default, removal is based on a least recently used (LRU) algorithm. The use of this algorithm ensures that no new data or recently accessed data is removed from cache. The process of deleting volumes in cache that were premigrated to tape is called migration. Volumes that were deleted in the cache and exist only on tape or cloud are called migrated volumes.
In a TS7720 and TS7760 (disk-only) configuration, no migrated volumes exist because there is no physical tape attachment. Instead, logical volumes are maintained in disk until they expire. For this reason, cache capacity for the TS7720 and TS7760 is larger than the capacity for the TS7740.
When a TS7720 and TS7760 is a member of a multicluster hybrid grid, virtual volumes in the TS7720 and TS7760 cache can be automatically removed. This removal is done by using a Volume Removal Policy if another valid copy exists elsewhere in the grid. A TS7700 grid refers to two or more physically separate TS7700 clusters that are connected to one another by using a customer-supplied Internet Protocol network.
On the TS7740, TS7720T, TS7760T, and TS7760C, a previously migrated volume must be copied back from tape or cloud into the TVC to be accessed. It must be copied because the host cannot directly access the physical tapes or cloud. When the complete volume is copied back into the cache, the host can access the data. The process of copying data back from tape or cloud to the TVC is called recall.
Figure 1-2 shows IBM TS7740, IBM TS7720T, and TS7760 TVC processing.
Figure 1-2 TS7740, TS7720T, and TS7760T Tape Volume Cache processing
With a TS7720 VEB and TS7760 VEC (disk-only), the virtual volumes are accessed by the host within the TVC.
Figure 1-3 shows the IBM TS7720 and IBM TS7760 TVC processing.
Figure 1-3 TS7720 and TS7760 Tape Volume Cache processing
Another benefit of tape virtualization is the data replication functions. Two, three, four, five, six, seven, and eight IBM TS7700 tape products can be interconnected. The connections can be through one of the following sets of links:
•Two or Four 1-gigabit (Gb) Ethernet links (copper or shortwave (SW) fiber)
•Two or Four 10-gigabit per second (Gbps) Ethernet links (longwave (LW) fiber)
These sets of links form a multi-cluster grid configuration. Adapter types cannot be mixed in a cluster. They can vary within a grid, depending on your network infrastructure. Logical volume attributes and data are replicated across the clusters in a grid. Any data that is replicated between the clusters is accessible through any other cluster in the grid configuration. Through remote volume access, you can reach any virtual volume through any virtual device. You can reach volumes even if a replication has not been made.
Setting policies on the TS7700 defines where and when you have multiple copies of your data. You can also specify for certain kinds of data, such as test data, that you do not need a secondary or tertiary copy.
You can group clusters within a grid into families. Grouping enables the TS7700 to make improved decisions for tasks, such as replication or TVC selection.
Depending on the configuration, multiple TS7700 tape products that form a grid provide the following types of solutions:
•High availability (HA)
•Disaster recovery (DR)
•HA and DR
•Metro and global business continuance
Before R3.2, a multi-cluster grid configuration presented itself to the attached hosts as one large library with the following maximums:
•512 virtual devices for a two-cluster grid
•768 virtual tape devices for a three-cluster grid
•1024 virtual tape devices for a four-cluster grid
•1536 virtual devices for a six-cluster grid
These numbers can now be exceeded in steps of 16 virtual drives, up to 496 virtual devices per cluster.
The copying of the volumes in a grid configuration is handled by the clusters, and it is not apparent to the host. By intermixing TS7720, TS7740, and TS7760 Models you can build a hybrid two, three, four, five, six, seven, or eight cluster grid.
Figure 1-4 shows multiple IBM TS7700 tape products in an example of possible host and grid connections.
Figure 1-4 Multiple TS7700 tape products that depict possible host and grid connections
For TS7740, TS7720T, and TS7760T grid configuration, each TS7740, TS7720T, and TS7760T manages its own set of physical volumes. Each maintains the relationship between logical volumes and the physical volumes on which they are located. Similarly, the TS7760C maintains the relationship between the logical volumes and cloud object stores where they reside.
The clusters in a TS7700 grid can be, but do not need to be, geographically dispersed. In a multiple cluster grid configuration, two TS7700 clusters are often located within 100 kilometers (km) or 62 miles of each other, whereas the remaining clusters can be located more than 1000 km (621.37 miles) away.
Cloud storage tier attachment further extends this distance and capability. This configuration provides a highly available and redundant regional solution. It also provides a remote DR solution that is outside of the region. A multi-cluster grid supports the concurrent growth and reduction of cluster counts.
1.4 Benefits of tape virtualization
The current global marketplace is increasingly information-oriented, which has far-reaching implications for businesses. The ability to rapidly and securely access information can create a competitive advantage. The following information-related business trends are causing an explosion of information and complexity in data centers:
•Information availability requirements are increasing.
•Information security threats and privacy regulations are increasing.
•Information compliance is more complex, and penalties are more severe.
•Information retention periods are longer, often exceeding the life of the storage media.
IBM offers an extraordinary range of systems, storage, software, and services that are based on decades of innovation. This range is designed to help you get the best solutions for your business requirements. It also manages challenges, such as exploding data growth, new applications, dynamic workloads, and new regulations. IBM Information Infrastructure intelligently stores, retrieves, protects, and distributes information to help you get a competitive advantage. Converting data centers to service-oriented architectures (SOAs) helps you identify and support multiple service levels, including information services.
Certain information services must be high speed to support websites and databases. In some cases, information services must be multiplexed to multiple locations, or require extra encryption and overwrite protection. IBM Information Infrastructure helps you apply the correct services and service levels so that vital information can be delivered. IBM Information Infrastructure solutions are designed to help you manage this information explosion. They also address challenges of information compliance, availability, retention, and security.
This approach helps your company move toward improved productivity and reduced risk without driving up costs. The IBM TS7700 is part of the IBM Information Infrastructure. This strategy delivers information availability, supporting continuous and reliable access to data. It also delivers information retention, supporting responses to legal, regulatory, or investigatory inquiries for information.
The TS7700 can be the answer to the following challenges:
•Enterprise storage platform to support business in the cloud era
•Growing storage requirements
•Shrinking backup windows
•The need for continuous access to data
You can expect the following types of benefits from tape virtualization:
•Brings efficiency to the tape operation environment
•Reduces the batch window
•Provides HA and DR configurations
•Provides fast access to data through caching on disk
•Provides optional use of current tape drive, tape media, and tape automation technology
•Provides optional use of filling high capacity media to 100%
•Provides many tape drives for concurrent use
•Provides data consolidation, protection, and sharing
•Requires no additional software
•Reduces the total cost of ownership
1.5 Managing the IBM TS7700
The TS7700 uses a Management Interface (MI) to manage key management functions:
•Grid configuration
•Logical and volume cartridge management
•Constructs management
•Monitoring and utilization overview
•Monitoring and defining partitions and premigration queues
•Ownership takeover mode
•User access and roles management
•Stand-alone volume mount support
•Pool encryption setting modification
•Library Request Command panel
•Cloud Storage Tier definitions and management
The TS7700 also includes a set of commands and enhanced statistical reporting.
1.6 Data storage values
The IBM TS7700 documentation displays data storage values that use decimal (base-10) prefixes and binary (base-2) units of measurement. Decimal units, such as kilobytes (KB), megabytes (MB), GB, and TB, are commonly used to express certain values. However, the base of the units can be misleading.
To prevent confusion, IBM uses a convention to differentiate between binary and decimal units. At the kilobyte level, the difference between decimal and binary units of measurement is relatively small (2.4%). This difference grows as data storage values increase. When values reach terabyte levels, the difference between decimal and binary units approaches 10%.
Both decimal and binary units are available throughout the TS7700 Tape Library documentation. Table 1-1 compares the names, symbols, and values of the binary and decimal units.
Table 1-1 Names, symbols, and values of the binary and decimal units
|
Decimal
|
Binary
|
||||
|
Name
|
Symbol
|
Value
(base-10)
|
Name
|
Symbol
|
Value
(base-2)
|
|
kilo
|
K
|
103
|
kibi
|
Ki
|
210
|
|
mega
|
M
|
106
|
mebi
|
Mi
|
220
|
|
giga
|
G
|
109
|
gibi
|
Gi
|
230
|
|
tera
|
T
|
1012
|
tebi
|
Ti
|
240
|
|
peta
|
P
|
1015
|
pebi
|
Pi
|
250
|
|
exa
|
E
|
1018
|
exbi
|
Ei
|
260
|
Table 1-2 shows the increasing percentage of difference between binary and decimal units.
Table 1-2 Increasing percentage of difference between binary and decimal units
|
Decimal value
|
Binary value
|
Percentage difference
|
|
100 kilobytes (KB)
|
97.65 kibibytes (KiB)
|
2.35%
|
|
100 megabytes (MB)
|
95.36 mebibytes (MiB)
|
4.64%
|
|
100 gigabytes (GB)
|
93.13 gibibytes (GiB)
|
6.87%
|
|
100 terabytes (TB)
|
90.94 tebibytes (TiB)
|
9.06%
|
|
100 petabytes (PB)
|
88.81 pebibytes (PiB)
|
11.19%
|
|
100 exabytes (EB)
|
86.73 exbibytes (EiB)
|
13.27%
|
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.