Professional Documents
Culture Documents
ibm.com/redbooks
Redpaper
International Technical Support Organization Tape Automation with IBM ^ xSeries Servers December 2001
Take Note! Before using this information and the product it supports, be sure to read the general information in Special notices on page 107.
First Edition (December 2001) This edition applies to tape subsystems supported by IBM for connection to IBM ^ xSeries servers. Comments may be addressed to: IBM Corporation, International Technical Support Organization Dept. HZ8 Building 662 P.O. Box 12195 Research Triangle Park, NC 27709-2195 When you send information to IBM, you grant IBM a non-exclusive right to use or distribute the information in any way it believes appropriate without incurring any obligation to you.
Copyright International Business Machines Corporation 2001. All rights reserved. Note to U.S Government Users - Documentation related to restricted rights - Use, duplication or disclosure is subject to restrictions set forth in GSA ADP Schedule Contract with IBM Corp.
iii
iv
Contents
Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .v Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii The team that wrote this Redpaper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Notice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii IBM trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii Chapter 1. Deciding on a tape library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Do I need a tape library? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Tape library selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.1 Tape library capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.2 Number of tape drives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 3 4 4 6
Chapter 2. Library Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.1 Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.1.1 Linear Tape Open . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.1.2 Super Digital Linear Tape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.2 IBM 3600 automation products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.2.1 IBM 3600 Series 900 GB/1.8 TB LTO Tape Autoloader . . . . . . . . . . . . . . . . . . . . 14 2.2.2 IBM 3600 Series 2/4 TB LTO Tape Library. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3 IBM 3502 DLT automation products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.3.1 IBM 280/560 GB DLT Tape Autoloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.3.2 IBM 490/980 GB DLT Tape Library. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.4 IBM 4mm automation product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.4.1 IBM 120/240 GB DDS/4 4mm Autoloader. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.5 Tape library comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.5.1 Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.5.2 Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.5.3 Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.5.4 Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.5.5 Form factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.5.6 Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.5.7 Additional features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.5.8 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.6 Supported solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Chapter 3. Software configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Tivoli Storage Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2 Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.3 Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.4 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.5 Other tape usages within Tivoli Storage Manager . . . . . . . . . . . . . . . . . . . . . . . . 3.2 VERITAS Backup Exec for Windows 2000 / NT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3 Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 31 34 38 41 50 66 67 68 71 76 v
3.2.4 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 CA BrightStor ARCserve 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chapter 4. Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Business recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 System recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Tivoli Disaster Recovery Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Bare Metal Restore for TSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.3 BrightStor ARCserve Disaster Recovery Option. . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.4 VERITAS Backup Exec Intelligent Disaster Recovery . . . . . . . . . . . . . . . . . . . . . 4.3.5 St. Bernard Open File Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.6 BrightStor ARCserve backup agent for open files. . . . . . . . . . . . . . . . . . . . . . . . . 4.3.7 VERITAS Backup Exec Open File Option. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Application recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Data recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix A. TSM element numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IBM 3600 Series 2/4 TB LTO Tape Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IBM 3600 Series 900 GB/1.8 TB LTO Tape Autoloader . . . . . . . . . . . . . . . . . . . . . . . . . . IBM 3502-x14 DLT Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IBM 3502-108 DLT Autoloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IBM 120/240 GB DDS/4 Autoloader. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Related publications . . . . . . . . . . . . . . . . . . . . . . . . . . . IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Referenced Web sites . . . . . . . . . . . . . . . . . . . . . . . . . . . How to get IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . IBM Redbooks collections. . . . . . . . . . . . . . . . . . . . . . ...... ...... ...... ...... ...... ...... ....... ....... ....... ....... ....... ....... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ......
79 82 83 86 89 93 93 94 94 95 95 97 97 97 98 98 98 99
101 102 103 103 104 104 105 105 105 105 106 106
vi
Preface
This Redpaper describes the current line of IBM xSeries tape automation products. These include the already existent 3502 DLT and DDS/4 series, as well as the new 3600 LTO series. We will discuss the need for tape products and libraries in particular, as well as give you library selection criteria. After describing the current products in details and positioning them to each other, we will provide information on how to use the tape libraries in conjunction with leading backup software. These include IBMs Tivoli Storage Manager, CA ARCserve and Veritas Backup Exec.
Frank Schallmoser is a level 2 support specialist for IBM-Intel based servers and has specialized in tape products and backup software. He holds a degree in applied Physics from the Technical High School of Aachen department Juelich (Germany) and the University of Coventry (UK). Before he joined IBM UK in 1996, he worked for an IT company in Germany, setting up hardware, software and administrating Novell and UNIX networks. He is a co-author of IBM Tape Solutions for Storage Area Networks and FICON, SG24-5474. Thanks to the following people for their contributions to this project: Steve Russell ITSO Specialist, xSeries, IBM ITSO Raleigh Center, IBM Gail Christensen ITSO Editor, IBM ITSO Raleigh Center, IBM Paul Chenger xSeries Technology Lab, IBM Raleigh John Gates xSeries Storage Development, IBM Raleigh
vii
Notice
This publication is intended to help storage and system administrators who are interested in or installing the current range of xSeries tape automation products.
IBM trademarks
The following terms are trademarks of the International Business Machines Corporation in the United States and/or other countries:
AFS AIX DB2 DFS e (logo) Enterprise Storage Server FICON IBM Informix Magstar MVS Netfinity Redbooks Redbooks Logo S/390 Sequent Tivoli WebSphere xSeries z/OS Lotus Notes Domino
Comments welcome
Your comments are important to us! We want our papers to be as helpful as possible. Send us your comments about this Redpaper or other Redbooks in one of the following ways: Use the online Contact us review Redbook form found at:
ibm.com/redbooks
viii
Chapter 1.
1.1 Introduction
Tape media has a long history of offering high capacity at low cost when it comes to storing data. In the current storage environment, however, tape is becoming increasingly under pressure from other technologies. Storage devices such as CD-ROM drives, optical drives and hard disks have become dramatically cheaper, and these have some potential advantages for delivering reduced hardware costs for a backup solution. For many users, however, tape continues to be the media of choice, since it delivers features that are not available in the alternatives. These include: Lowest cost per bit of storage Tape media has an estimated cost of less than $5 per GB. Highest capacity per unit Comparing tape media to CD-R shows a massively higher capacity. For example an LOT cartridge is capable of holding 100 GB of data. Stored on CD-R media, this would require a total of 170 CDs. In the past, it has been assumed that tape media are unstable and slow. Where this was certainly true a few years ago, new technologies have dramatically improved reliability and performance.
Uses for tape devices differ, but most of them can be placed in the Storage Management arena of an enterprise. Storage management includes several disciplines, including: Data protection Data archiving Space management Data protection is probably the best known and most often implemented part of storage management. It includes backing up data, providing insurance against data loss. In todays Intel server world there is a lot of information available on how to increase system reliability and availability. There are many solutions that help achieve this, including cluster solutions, RAID array solutions and Storage Area Networks (SAN). However, where these solutions provide assurance for system component failure, they do not protect against data loss due to application failures, human error or viruses. For this reason it is important that you have a backup strategy to provide the possibility to recover your data in the event of a failure. Backups not only secure your current data but also allow you the possibility to recover previous files that have been overwritten by newer versions. Data archiving is used to provide long term storage for important data. Most companies treat data from a legal point of view. Keeping that data in online storage such as disks is not the proper solution, due to high management costs. A better solution would be to place the data in offline storage, such as tapes. Important when implementing archive solutions is media stability. Tape media currently have an estimated life expectancy of 10 to 20 years, depending on the technology and storage conditions. Where this is enough for some documents, others might require longer retention periods. In that case, other data carriers like optical media, microfilm or even paper still need to be considered. A consideration for long-term storage is a factor called technology obsolescence. This indicates the possible problems that might arise when you are required to read a tape that has been created 20 years ago. Will you still have the required hardware available? Will it still work? Short-term archiving on the other hand is much more common, and ideally suited for low-cost storage. Imagine the case where you want to keep project data for a couple of years, but do not want to place it in online storage. In that case, tape media are ideal, especially when combined with archiving management software. Another, less common use of tape media is space management. The main purpose of space management is similar to short-term archiving of data. Both move unused or infrequently used data from disk storage to tape. The main difference is how moved data is handled. In the archiving solution, retrieving the data involves interaction with the storage manager. The user will need to request a retrieve of his archive data in order to consult it. With space management, this process becomes transparent to the user. When he requests his data, the space management software will automatically detect that it concerns migrated data. Next, the software will automatically recall the data from tape storage back to disk, where it can be consulted. The only inconvenience for the user requesting the data will be that its a bit slower to get his data. That is why most space management software use certain rules, for example the last access date, before moving data. Space management has been used for years on mainframe systems, such as MVS. In the open world, it has become more and more popular. When considering tape devices as secondary storage for space management, the most important factor is the data access time. This will typically require smaller capacity, but faster tape technology.
role. If you require capacity that is larger than the one that fits on a single data cartridge, you will need two stand-alone drives. If you then need to double your performance, the resulting number of single tape devices is four. The same performance and capacity combination can be obtained by selecting a two-drive tape library, leading to a less expensive solution. A final but very important factor is centralization. As discussed above, tape libraries are the ideal solution for implementations that require high-capacity, high-performance backup devices. This allows to consolidate stand-alone tape backup solutions to a central backup server using a tape library. Advantages include lower operational and maintenance costs, better control and less need for backup skills. Instead of having the system administrators handle their own backup, one central service can manage the entire backup process. One of the disadvantages of a tape library compared to a single drive is its complexity in setup and management. Managing a tape library in an efficient way will call for more advanced software than that available in the operating system. However, these products typically add functions that simplify backup and tape management, making them a requirement for medium to large backup implementations.
The above table leads us to the following results: When you do a daily full backup, you will need to have 12 slots for a weekly change of the tapes. When doing a full/incremental backup, you only need 7 slots. This means that a 14 slot library could hold two weeks of backup. A final step in this example is to take the cleaning cartridge into account. This cartridge will take up one slot, which needs to be added to the total number of tape slots.
This completes the introduction to the need and selection of a tape library. For additional information, please refer to the following publications: Netfinity Tape Solutions, SG24-5218 The IBM LTO Ultrium Tape Libraries Guide, SG24-5946 IBM Tape Solutions for Storage Area Networks and FICON, SG24-5474 Implementing IBM LTO Tape in Linux and Windows, SG24-6268 Using IBM LTO Ultrium with Open Systems, SG24-6502 IBM Magstar Tape Products Family: A Practical Guide, SG24-4632-03
Chapter 2.
Library Overview
This chapter discusses the features of tape automation products. These include: Tape drive technology Library capacity Performance information Interfaces Additional features Product comparison Supported solutions In addition, this chapter provides an overview of the current IBM tape automation solutions for IBM xSeries servers. We also describe the hardware installation and configuration. Specifically, the following tape libraries are covered: IBM 3600 Series 900 GB/1.8 TB LTO Tape Autoloader on page 14 IBM 3600 Series 2/4 TB LTO Tape Library on page 15 IBM 280/560 GB DLT Tape Autoloader on page 19 IBM 490/980 GB DLT Tape Library on page 20 IBM 120/240 GB DDS/4 4mm Autoloader on page 21
2.1 Technology
As we see more and more tape products appear on the market, some of them using new technologies, it is important to know about these products and understand what each offers in terms of performance, capacity, growth, investment protection and connectivity. This section provides an overview of the technologies that are available from the tape automation products for IBM xSeries range. We discuss the newer tape technologies, mainly the Linear Tape Open (LTO) and the Super Digital Linear Tape (SDLT). For other existing technologies, for example Digital Linear Tape (DLT) and Digital Audio Tape (DAT), see Netfinity Tape Solutions, SG24-5216.
Note: Throughout this document we will no longer refer to DAT but instead to Digital Data Storage Four (DDS/4). This is the recording format that is currently used on DAT drives and as such indicates the standard of the product.
10
Table 2-1 provides an overview of possible future enhancements to the Ultrium format.
Table 2-1 LTO Ultrium roadmap Generation 11 Capacity Transfer rate2 Recording Format Media 100 GB 10-20 MBps3 RLL 1,7 MP Generation 2 200 GB 20-40 MBps PRML MP Generation 3 400 GB 40-80 MBps PRML MP Generation 4 800 GB 80-160 MBps PRML Thin Film
Notes: 1. The current products are first-generation products. 2. The figures given are native capacity and transfer rate. Compression usually doubles these figures, giving a 200 GB capacity and a 20-40 MBps transfer rate for first-generation Ultrium tapes. 3. The LTO standard allows transfer speeds ranging from 10 to 20 MBps in native mode. The currently available products from IBM typically perform at 15 MBps.
Note: All the capacity and performance figures given for the LTO Ultrium and Accelis formats suppose the use of 8-head devices. The LTO standard also supports 4-head devices, which are less common on the market. The consequence of using only 4 heads instead of 8 is that the capacity and performance figures must be divided by two.
Accelis
The Accelis tape format is the implementation of LTO technology optimized for fast access to data. It uses a two-reel cartridge that loads at the middle of the tape to minimize access time. The Accelis tape format is targeted at automated environments and can enable a wide range of online data inquiry and retrieval applications. Data transfer speed is comparable to the Ultrium format. At this point, no devices are available that use the Accelis tape format. The first generation of Accelis allows 25 GB native capacity. A self-enclosed tape path in the cartridge eliminates tape threading, which greatly improves time to first data byte. Cartridges are loaded in the middle of the tape rather than at the beginning, reducing average search time for random files. Accelis provides for data transfer rates of 10-20 MBps in the first generation. Accelis is ideal for library use, with high-speed access to relatively short files. Accelis is suited for applications such as data mining and image retrieval, as well as traditional backup/restore. Cartridge memory (LTO-CM) enhances functionality by providing a non-contacting passive radio frequency (RF) interface embedded in the cartridge, allowing information retrieval (calibration information, manufacturers' data and information about initialization). The advantage is that this information can be retrieved without having to load the tape. Data integrity features include two levels of error correction that can provide recovery from longitudinal media scratches. Read-While-Write (RWW) capability allows real-time verification of written data.
11
Table 2-2 LTO Accelis roadmap Generation 1 Capacity1 Transfer rate1 Recording Format Media Data access time 25 GB 10-20 MBps RLL 1,7 MP < 10 sec Generation 2 100 GB 20-40 MBps PRML MP < 8 sec Generation 3 200 GB 40-80 MBps PRML MP < 7 sec Generation 4 400 GB 80-160 MBps PRML Thin Film < 7 sec
Notes: 1. The figures given are native capacity and transfer rate. Compression usually doubles these figures, giving a 200 GB capacity and 20-40 MBps transfer rate for first-generation Ultrium tapes.
For more information on the LTO products and technology, see: http://www.storage.ibm.com/hardsoft/tape/lto/lto.htm http://www.lto.org
12
New buckler design Where the buckler design of the DLT IV tapes was a regular source of tape breakdown the SDLT tapes provide a new buckling mechanism that is more reliable. See: http://www.quantum.com/NR/rdonlyres/000001c0lgzhcytizmgzqqio/wptapehandling.pdf As with LTO formats, Quantum has designed a roadmap for its SDLT product line. Table 2-3 provides an overview of the specifications of future SDLT generations.
Table 2-3 SDLT product roadmap Generation 1 SDLT 2201 Capacity2 Transfer Rate2 Media Availability date 110 GB 11 MBps SDLT I 2001 SDLT 320 320 GB 16 MBps SDLT I 2002 Generation 2 SDLT 640 640 GB 32 MBps SDLT II 2004 Generation 3 SDLT 1280 1280 GB 50 MBps SDLT III 2005 Generation 4 SDLT 2400 2400 GB 100 MBps SDLT IV 2007
Note: 1. This is the currently available product. 2. The figures given are native capacity and transfer rate. Compression usually doubles these figures, giving a 220 GB capacity and a 22 MBps transfer rate for SDLT 220 devices.
For more information on SDLT technology see: http://www.dlttape.com/Technology/Default.htm http://www.quantum.com/Products/Quantum+l+DLTtape/SDLT+220/Default.htm Currently, IBM does not offer a tape automation product that uses the SDLT 220 drives, but rather the SDLT 220 single tape drive. For more information, see: http://www.pc.ibm.com/ww/eserver/xseries/tape.html
13
Table 2-4 Overview of 3600 LTO automation products Model Number Number of drives Maximum number of cartridges Native capacity1 Connection type Optional Fibre Channel connection Expandable 3600 LTO autoloader 3600-109 1 LTO tape drive 9 900 GB Ultra-2 LVD SCSI NO NO 3600 LTO tape library 3600-220 (tower) 1 LTO tape drive (2 maximum) 20 2 TB Ultra-2 LVD SCSI YES FC-AL interface NO 3600 LTO tape library 3600-R20 (rack) 1 LTO tape drive (2 maximum) 20 2 TB Ultra-2 LVD SCSI YES FC-AL interface YES with up to two 3600-LXU units.
Note: 1. The total capacity of the library can double using compression.
The autoloader has the following features: Up to 900 GB total native capacity, 1.8 TB using 2:1 compression. Data transfer rate of 15 MBps native, 30 MBps with compression. A six-slot removable magazine and three fixed slots in the back of the autoloader. The autoloader is primarily designed as a tower. It can be installed in a rack on a fixed shelf. Each unit occupies 6U of rack space. Integrated barcode reader for improved media management. Front -anel LCD display and operator control panel. Support for industry-leading backup and restore application software.
14
The 3600 autoloader option kit includes the following items: The LTO tape autoloader assembly. A country-specific power cord. An external 4-meter LVD SCSI cable. Publications, including: Online version of the user manual. Quick Install Guide (English). One data cartridge. One cleaning cartridge. An external 1-meter LVD SCSI jumper cable. Barcode kit. An LVD external active SCSI terminator. Trial version of the VERITAS Backup Exec software. Trial version of the Computer Associates ARCserve software. Figure 2-2 provides an overview of the required connections for the 3600-109.
SCSI Terminator
To Host
Figure 2-2 Connecting the 3600-109
SCSI Cable
15
Figure 2-3 The IBM 3600 Series 2/4 TB LTO Tape Library
The library has the following features: Up to 2 TB total native capacity native, 4 TB using 2:1 compression. A data transfer rate of 15 MBps native, 30 MBps with compression. Two drawers, each with two removable magazines that both hold five cartridges. The library comes in two form factors, as a tower model or a rack model that occupies 5U of rack space. Integrated barcode reader for improved media management. Front-panel LCD display and operator control panel. A remote management card is included to provide remote manageability and diagnostics. Hot pluggable drives. Support for industry-leading backup/restore application software. Modular building block architecture for the tower model using the 2-Drive, 20-Cartridge Expander Module. Up to two tape drives using the 3600 Series LTO Drive Upgrade Option. Native Fibre Channel support using the IBM Fibre Tape Automation Adapter. The 3600 Series 2/4 TB LTO tape library option kit includes the following items: LTO tape library assembly. Country-specific power cord. An external 4-meter LVD SCSI cable. Publications, including: Online version of the user manual. Quick Install Guide (English). One data cartridge. One cleaning cartridge. External 1-meter LVD SCSI jumper cable. Barcode kit. An LVD external active SCSI terminator. Trial version of the VERITAS Backup Exec software. Trial version of the Computer Associates ARCserve software. Figure 2-4 provides an overview of the required connections for the 3600-x20:
16
SCSI Terminator
17
LAN
T
As shown in Figure 2-5, each expander module requires a separate SCSI connection to the host system. The reason for this requirement is that due to the speed of the LTO tape drives, only two devices must be connected on the same SCSI bus. This is mainly to avoid saturation of the SCSI bus. When you add expander modules without additional tape devices, there is no need to add SCSI connections to the host adapter. Figure 2-5 also shows that the LTO 3600 tape library can be connected to the customers LAN; this is for systems management purposes. The systems management card enables remote management and diagnostics of the tape library via a Web browser.
18
SCSI Terminator
LAN
T T
With the fiber card now installed you have native fiber connectivity. We will discuss the possible connections to the SAN or host adapters (see Connectivity on page 26).
19
The IBM 3502-108 DLT Tape Autoloader has the following features: Up to 280 GB total capacity native, 560 GB with compression. Data transfer rate of 5 MBps native, 10 MBps using compression. Removable 6-cartridge magazine plus 2 fixed slots. The autoloader is primarily designed as a tower. It can be installed in a rack, however, using a fixed shelf. Front-panel LCD display and operator control panel. Support for industry-leading backup/restore application software. The 3502-108 DLT Tape Autoloader option kit includes the following items: The autoloader assembly. Country-specific power cord. A 3-meter SCSI cable. An 2940U2B SCSI adapter. User Manual. DLT data cartridge. DLT cleaning cartridge. 0.8 mm. (68-pin adapter) A 0.5 meter SCSI jumper cable. A 68-pin SCSI terminator. A trial version of VERITAS Backup Exec software. A trial version of Computer Associates ARCserve.
20
The IBM 3502-314/R14 Tape Library has the following features: Up to 560 GB total capacity native, 980 GB using compression. Data transfer rate of 5 MBps native, 10 MBps using compression. Two removable 7-cartridge magazines. The library comes in two form factors, as a tower and a rack that occupies 4U of rack space. Integrated barcode reader for improved media management. Front-panel LCD display and operator control panel. Support for a wide variety of backup/restore software. The 3502-314/R14 DLT Tape Library option kit includes the following items: The library assembly. Country-specific power cord. A 3-meter SCSI cable. A 2940U2B SCSI adapter. User manual. DLT data cartridge. DLT cleaning cartridge. 0.8-mm (68-pin adapter) A 0.5 meter SCSI jumper cable. A 68-pin SCSI terminator. A trial version of VERITAS Backup Exec software. A trial version of Computer Associates ARCserve.
21
Table 2-6 Overview of DDS/4 autoloader Model number Number of drives Max. number of cartridges Maximum capacity (native / compressed) Connection type Upgradable 120/240 GB DDS/4 tape autoloader 1 DDS/4 tape drive 6 120 GB / 240 GB Ultra -2 LVD SCSI No 120/240 GB DDS/4 tape autoloader with 3503-B1X 1 DDS/4 tape drive 6 120 GB / 240 GB Ultra -2 LVD SCSI No
The IBM 120/240 GB DDS/4 Tape Autoloader has the following features: Up to 120 GB total capacity native, 240 GB using compression. Data transfer rate of 3 MBps native, 6MBps using compression. A six-cartridge removable magazine. Internal installation in any standard 5.25-inch full-height bay. Front-panel LCD display and operator control panel. Support for industry-leading backup/restore application software. The 120/240 GB DDS/4 Tape Autoloader option kit includes the following items: The 120/240 GB DDS/4 Tape Autoloader A user manual. Mounting screws. Five data cartridges. A cleaning cartridge. Two cartridge magazines. A trial version of VERITAS Backup Exec software. A trial version of Computer Associates ARCserve.
22
2.5.1 Performance
The performance of a tape library is normally defined using two parameters: The performance of the tape technology used within the library. The number of tape devices that can be installed in a library. To obtain the overall potential data throughput of a tape library, the throughput of the tape drives should be multiplied by the number of tapes available. Table 2-7 provides an overview of the total library data throughput. The data throughput rates provided describe the native speed.
Table 2-7 Library throughput Number of drives 3600 LTO tape library (tower) Standard Maximum 3600 LTO tape library (rack) Standard Maximum 3600 Series tape autoloader 3502 DLT tape library Standard Maximum 3502 DLT tape autoloader DDS/4 tape autoloader 1 2 1 61 1 1 3 1 1 Drive performance 15 MBps 15 MBps 15 MBps 15 MBps 15 MBps 5 MBps 5 MBps 5 MBps 3 MBps Library throughput 15 MBps 30 MBps 15 MBps 90 MBps 15 MBps 5 MBps 15 MBps 5 MBps 3 MBps
Note: 1. The total of 6 drives can be obtained using 2 3600-LXU expander units.
Where the throughput is definitely an indication of the possible transfer speed of a tape drive or library, there are some considerations that need to be taken. When comparing products that will operate in a given condition, the throughput is a valid comparison tool. It is, however, not an indication of what throughput speeds you will obtain in the final setup. Several other factors play a role, including: Search time required to locate data and position a tape. Type of data transferred. Tape devices are typically optimized for data streaming, which means that data will flow continuously from or to the tape drive. If the used environment is characterized by a large number of small files, this streaming will probably not take place. The resulting transfer rate will be smaller. This becomes increasingly more important when trying to use the total throughput of the library. In the case of a fully configured 3600 LTO library for example, the 90 MBps rate can only be obtained if the host is able to stream the data to all six drives at the same time. If multiple tape mounts are required, the time required to mount a tape in a drive becomes an important value. This value is a function of the following points: Time required to locate a volume and position the picker. Time to move volume from the slot to the tape devices. Time required to load and position the tape volume in the tape devices.
23
2.5.2 Capacity
The capacity of a autoloader or library is dictated by the number of cartridges that it can hold and the tape technology used. As with the throughput values, the total capacity can be calculated by multiplying the number of available slots by the capacity of the cartridges used. Table 2-8 provides an overview of the total library capacity. The values provided describe the native capacity.
Table 2-8 Library capacity Number of slots 3600 LTO tape library (tower) 3600 LTO tape library (rack) 3600 LTO tape library (rack) with two 3600-LXU units 3600 Series tape autoloader 3502 DLT tape library 3502 DLT tape autoloader DDS/4 tape autoloader 20 20 60 9 14 8 6 Cartridge capacity 100 GB 100 GB 100 GB 100 GB 35 GB 35 GB 20 GB Library capacity 2 TB 2 TB 6 TB 900 GB 490 GB 280 GB 120 GB
As with the data throughput values, there are some considerations that are important: Since tape libraries typically operate in a low-maintenance environment, you should add a cleaning tape to the library. This means that the number of available slots should bbe lowered by the number of cleaning cartridges (typically 1). In some situations, you will not use the entire capacity of a cartridge. For example, when you want to group all backup data of one client system on a cartridge, you could end up with a large amount of cartridges that are not fully used. Therefore, you should understand the use of your library, and base the decision about capacity on either the total data capacity or the number of available slots.
2.5.3 Compression
Where compression is not really a tape library factor, but more dependent on the tape technology used, it is commonly used when describing library capacity and performance. The current available tape product in the xSeries range use a compression algorithm that will typically double the capacity and performance. Care should be taken, however, when taking this compression factor into account. This because it strongly depends on the type of data that is written to the tape. Hardware tape compression is basically the same as normal software compression. The only difference is that hardware compression will be performed by the tapes firmware, rather than by the host sending the data. The result is that files that are already compressed at the origin will not gain from hardware compression.
2.5.4 Scalability
When selecting a library, a customer is always faced with the problem of data growth. Since the capacity and performance needs will be largely based on the amount of data that needs to be placed on tape media, having a valid figure is very important. However, due to recent surges in data use, determining future needs has become increasingly difficult. A possible solution is to invest in a library that largely exceeds current needs, but is capable of coping with the projected data growth. 24
Tape Automation with IBM ^ xSeries Servers
The problem with this decision is that the investment required is much higher. Another, more reasonable solution would be to choose a library that can grow over time, meaning that there is the possibility to add tape devices (increasing performance) or capacity. Table 2-9 shows the expansion possibilities available for the current libraries.
Table 2-9 Library expansion possibilities Capacity expansion 3600 LTO tape library (tower) 3600 LTO tape library (rack) 3600 Series tape autoloader 3502 DLT tape library 3502 DLT tape autoloader DDS/4 tape autoloader No Yes No No No No Performance Expansion Yes Yes No Yes No No
25
2.5.6 Connectivity
The possibilities we have connecting the library to the host system is called connectivity. Besides the standard SCSI connectivity, SAN attachment is becoming more and more important. There are many reasons for this, ranging from offloading network traffic to a SAN with LAN free solutions, sharing the backup device between servers (tape pooling) or distant disaster recovery solutions, to backing up data off ite or to and from other locations. SAN connectivity through the Fibre Channel protocol can be obtained using two methods: Direct attachment of the devices. Attachment through separate components, called gateways or routers. Table 2-11 provides an overview of the connectivity options for the different libraries.
Table 2-11 Library connectivity SCSI interface 3600 LTO tape library 3600 Series tape autoloader 3502 DLT tape library 3502 DLT tape autoloader DDS/4 tape autoloader LVD LVD SE SE LVD SAN Connectivity Fibre Channel Tape Automation adapter Not supported IBM SAN Data Gateway SCSI Tape Router (2108-R03 and Feature Code 2840) IBM SAN Data Gateway SCSI Tape Router (2108-R03 and Feature Code 2840) Not supported
The following two diagrams provide an overview of connectivity options using Fibre Channel connections. Figure 2-10 shows the connectivity using the integrated Fibre Channel connections available on the 3600 Series tape libraries. Connections can be made using point-to-point mode, or libraries can be directly attached to the SAN fabric.
26
For tape devices that dont support native fiber connections, you will need additional hardware to provide fiber connectivity. For this, IBM offers a range of SAN data gateway products. These products will offer some investment protection for customers who dont wish to purchase new tape products, but require some of the functionality that they provide.
I/O ports are predefined slots in a library where volumes are moved when check-in or checkout operations occur. The advantage is that the library does not need to be opened to handle these volumes, ensuring continuity of operations. Other names for I/O slots include entry/exit ports and mail slots. Currently, only the 3600 library (3600-220 and R20) have this functionality.
2.5.8 Summary
Figure 2-12 on page 28 shows each product and its relative position in comparison to the other automated solutions available for IBM xSeries servers. With the range of products available, there is always going to be an overlap. Lets look at an example. Although the performance of the 3600 tape autoloader and the 3600 tape library are the same in terms of
elbaC ISCS =
yawetaG ataD
yawetaG ataD
noi t ar u gifnoC noi tcennoC tceriD noit aru gifnoC hctiw S & bu H
revres seireSx revres seireSx
27
throughput, the 3600 tape library has the capacity to hold an additional tape or additional tapes if using the Library Expander Module (LXU), which is available for the rack mount library, to provide greater throughput. This is one reason that the library has been positioned above the autoloader. Another reason would be capacity. The figure also shows the market position of each product but this will change as customer requirements change, especially as storage requirements continue to grow, and new products are announced. Here we are only referring to those products offered as options by IBM xSeries servers and how IBM markets those products within this range. One other point to note: when we previously described some of the machines as entry-level products we didnt necessarily mean that they are low-end solutions. For example, the LTO autoloader is an entry-level solution with regards to LTO tape products, but is not a low-end solution. Entry-level products are generally less complex machines; therefore they make excellent first-time automation products for customers who require greater capacity than a single tape drive can provide.
Throughput
High
Medium
3600-109
3600-220/R20
Low
3502-314/R14 3502-108
DDS/4 Autoloader
Capacity
Figure 2-12 Product positioning
Table 2-12 provides a summary of all the topics discussed in the above sections.
28
Table 2-12 Product comparison SCSI interface Form factor Data cartridges (max) 20 603,4 9 14 14 6 5 No. of drives (std./max ) 1/2 1/63,4 1/1 1/3 1/3 1/1 1/1 Capacity (native/ compressed) 2/4 TB 6/12 TB 3,4 900 GB/1.8 TB 490/980 GB 490/980 GB 280/560GB 120/240 GB Backup Rate1 (native/ compressed) 15/30 MBps 15/30 MBps 15/30 MBps 5/10 MBps 5/10 MBps 5/10 MBps 3/6 MBps
3600 Series 2/4TB LTO Tape Library (Tower) 3600 Series 2/4TB LTO Tape Library (Rack) 3600 Series 900GB/1.8TB LTO Tape Autoloader 3502-314 DLT Tape Library (Tower) 3502-R14 DLT Tape Library (Rack) 3502-108 DLT Tape Autoloader 120/240 GB DDS/4 Tape Autoloader Notes:
1. Transfer rates are for single SCSI channel configurations. Tape libraries utilizing split library or dual host configurations may obtain higher rates. Data compression typically provides a 2X improvement in capacity and transfer rate, bur since data compression is affected by many factors, actual improvements may be more or less than 2X. 2. This is the standard connection. However, there is a fiber adapter option. This adapter installs in a 3600 Series Tape Library to allow native fiber connectivity. Each adapter supports up to two LTO drives. 3. Maximum configuration includes two 3600 Series 2-Drive, 20-Cartridge Expander Modules and additional LTO tape drive options. 4. The 3600 Series 2-Drive, 20-Cartridge Expander Module must be installed by IBM service. This installation service is included without additional charge. It is supported only with the 3600 Series LTO Tape Library (Rack) P/N 21P99xx. One additional EIA space has to be allowed when installing either one or two units (maximum) to accommodate a filler plate for cable routing. Up to two 3600 Series LTO Drive Upgrade Options can be installed in each module or the module can operate off the LTO drives installed in the LTO tape library.
29
30
Chapter 3.
Software configuration
This chapter discusses the installation and configuration of additional software products that can be used to manage your backup and restore operations and control your library. The following software products are discussed: Tivoli Storage Manager on page 31 VERITAS Backup Exec for Windows 2000 / NT on page 67 CA BrightStor ARCserve 2000 on page 82
31
TSM Server
Archive Retrieve Backup Restore
Tivoli Storage Manager manages your storage media based on a set of policies, which you can customize for different types of data and groups of users. You can also automate some operations, such as client backups and server database backups. TSM provides a wide range of versions for different platforms. Server platforms include Windows NT and 2000, AIX, Sun Solaris, HP UNIX and MVS. Client platforms include Windows platforms, Linux, Novell NetWare, AIX, HP UNIX and Sun Solaris. For details on other supported client platforms, please refer to: http://www.tivoli.com/support/storage_mgr/requirements.html The Tivoli Storage Management product set includes the following: Tivoli Storage Manager server The TSM server is the central component of the Tivoli Storage Management suite. The TSM server is responsible for storing and managing all client data. The server consists of the following three major components: Server application code The server database The TSM server database will be the heart of the entire TSM installation. All information about configuration and stored client data will be centralized in this database. The database is a relational database (proprietary), which has all major functions supported by relational database systems. This includes transaction logging, backup, point-in-time recovery and roll-forward capabilities. In addition, the database supports access through ODBC drivers (read-only) for reporting purposes. The servers storage pools Whereas the TSM server database stores all the meta or control data, the TSM server will make use of so-called storage pools to store the actual client data. These storage pools are made up of physical media, which can range from disks to tapes to optical devices.
32
Administrative interfaces The administrative interface component is used to configure and administer the TSM server. TSM provides three types of administrative interfaces: The Web administrative interface The command line interface (CLI) A plug-in into the Microsoft Management Console (MMC) interface These interfaces can be used locally (on the TSM server), as well as on remotely network connected workstations. The Web administrative interface has the advantage that no code is required on the workstation from where the TSM server is controlled. The CLI is only available when the code is installed. The CLI code is part of the TSM client installation package. TSM Backup/Archive client The Backup/Archive client component is installed on the TSM client systems. Its main purpose is to provide backup, restore, archive and retrieve functions. The TSM client is composed of the following components: Backup/Archive GUI interface, allowing user-controlled backups, restores, archives and retrieve. Command line interface, which can be used for batch processing. A scheduler service, allowing automation of all client functions. A Web Client Interface. This is a Java-based application that allows you to remotely run the Backup/Archive GUI clients. Tivoli Disaster Recovery Manager (DRM) This product enables you to create a disaster recovery plan, which can be used to set up a recovery plan for both the TSM server as well as the TSM clients. Note: The DRM component provides a tool to automatically create TSM client and TSM server recovery plans, based on the information stored on the TSM server. It does not provide a tool for complete system recovery. If this function is required, TSM proposes a third-party add-on, called Bare Metal Restore for TSM. For more information, please see: http://www.tkg.com/products/baremetal/tivoli/baremetal_tivoli.php For more information concerning DRM, please see: http://www.tivoli.com/products/index/disaster_recover_mgr Tivoli Space Manager Tivoli Space Manager uses hierarchical storage management (HSM) to automatically and transparently migrate infrequently accessed files to the TSM server while more frequently used files remain in local file systems for fast access. When a user requires access to a migrated file, it will be automatically recalled from the TSM servers storage. Tivoli provides a space management client for several UNIX platforms, and provides support for Windows NT and 2000 space management clients in combination with a product called OTG DiskXtender. For more information on OTG DiskXtender, see: http://www.otg.com/tivoli/
33
Tivoli Data Protection for application clients These products enable you to back up and restore specific applications. Currently, the following applications can be backed up through these application clients: Tivoli Data Protection for IBM Enterprise Storage Server (ESS) Tivoli Data Protection for EMC Symmetrix Tivoli Data Protection for NDMP Tivoli Data Protection for WebSphere Tivoli Data Protection for Informix Tivoli Data Protection for Lotus Domino (including the S/390 edition) Tivoli Data Protection for Microsoft Exchange Server Tivoli Data Protection for Microsoft SQL Server Tivoli Data Protection for Oracle Tivoli Data Protection for R/3
3.1.1 Concepts
All client data that is managed by TSM is stored in the TSM storage repository. This repository can consist of different storage devices, such as disk, tape, or optical devices, and controlled by the TSM server. To do so, TSM uses its own model of storage to view, classify, and control these storage devices, and to implement its storage management functionality. The main difference between the storage management approach of TSM and other commonly used systems is that the TSM concentrates on managing data objects instead of managing and controlling backup tapes. Data objects can be sub-file components (a sub-file is part of an entire file), files, directories or raw logical volumes that are backed up from the client systems. They can be objects such as tables or records from database applications, or simply a block of data that a client system wants to store on the servers storage. To store these data objects on storage devices and to implement storage management functions, TSM has defined some logical entities to classify the available storage resources. Most important is the logical entity called a storage pool. A storage pool describes a storage resource for one single type of media, such as a disk partition or a set of tape cartridges. Storage pools are the place where data objects are stored. A storage pool is made up of one or more storage pool volumes. For example, in the case of a tape storage pool, this would be a single physical tape cartridge. To describe how TSM can access those physical volumes to place the data objects on them, TSM has another logical entity called a device class. A device class is connected to a storage pool and specifies how volumes of this storage pool can be accessed. TSM organizes storage pools in one or more hierarchical structures. This storage hierarchy can span multiple server instances, and is used to implement management functions to migrate data objects automatically completely transparent to the client from one storage hierarchy level to another; or in other words, from one storage device to another. This function may be used, for example, to cache backup data (for performance reasons) onto a TSM server disk space before moving the data to tape cartridges. The actual location of all data objects is automatically tracked within the server database. Another important storage management function implemented within the TSM server is the ability to copy data objects asynchronously and to store them in different storage pools or on different storage devices, either locally at the same server system or remotely on another server system. It is especially important for disaster recovery reasons to have in the event of losing any storage media or the whole storage repository a second copy of data available somewhere in a secure place. This function is fully transparent to the client, and can be performed automatically within the TSM server. 34
Tape Automation with IBM ^ xSeries Servers
The storage pools themselves are defined in a so-called storage hierarchy. This allows TSM to set up a chain of storage pools that contain the clients data. The storage pools hierarchy is mainly used as follows: 1. A clients data will initially go to a disk storage pool (defined in the client's data copy group). The advantage of writing data first to disk and then to tape is not mainly due to performance reasons (however, with some older tape technology and the required time for tape mounts in a library this can play a role), but to limit the amount of required tape drives. A tape device is typically a single session storage device. This means that only one process can write data to the tape at one moment in time. Since this process is the backup or archive process in TSM, this would mean that if concurrent backup or archive processes for different clients would be required, the number of required tape devices would be equal to the number of these concurrent client processes. A disk device, however, is multi-session, meaning that different clients can write data to the same disk device at once. So, by defining a disk storage pool as the initial storage pool, you are allowing multiple client sessions independent of the number of tape devices available. 2. Since the initial disk storage pool cannot be created as big as required to hold all the clients data, the data should be moved to another storage pool, which will typically be a tape storage pool. This movement between storage pools is called migration. Migration is controlled by setting a high and low threshold on the lower storage pool. Whenever the occupancy of this lower storage pool goes above this high threshold, the TSM server will automatically start a migration, and move data to the next storage pool. This movement will continue until the low-level threshold is reached. This migration process is totally independent of the client processes, and is therefore considered asynchronous to the initial data movement from client to server. Some additional considerations on using tape storage with TSM are due to the fact that TSM does not work with so-called backup sets or tape rotation schemes, nor does it call for an intermittent full backup. TSM will manage the clients data entirely by itself based on the defined policies. This means that the tapes used within TSM don't have a predefined life. A consequence of this is that good TSM design calls for availability of all tapes. This allows you to do a restore of any version of a file without user intervention. This means that you need to have enough storage slots within the tape library. A second consideration is that since all tapes remain on site (in the tape library or close to it), there is no off-site copy in case of a disaster. To provide this possibility, TSM uses copy storage pools. Copy storage pools are similar to primary storage pools, but contain a copy of the data in the primary storage pools. So, these tapes can be stored in an off-site location, and can be used in case of a disaster. A second more common use of copy tapes is when a tape in a primary storage pool becomes defective. Figure 3-2 shows an overview of the main data movement in a TSM environment:
35
TSM Client
Backup Data
TSM Server Migration Tape Storage Pool Backup Stgpool Storage Hierarchy
The definition of a tape storage pool making use of a tape library is as follows (details on how to do this will follow later): 1. Define the tape library that you will use, giving it a logical device name by which it will be known to TSM (for example: IBM_3600). 2. Define the type of tape devices located in the tape library you just defined. This is called a Device Class. As with the library, the device class you define will need a name, which is user defined. Besides the type of devices, you can also define additional physical parameters, such as the number of mount points (a mount point is a logical drive associated with a physical drive; the number of mount points for a device class is determined by the mount limit for that class) and the tape format used. It is possible to define more than one device class within one tape library. This might be useful when using a library device that supports more than one type of tape devices within a library. 3. Define the tape drives located within the library defined in step 1. Again, you should name the tape drive. 4. Define a storage pool that will use the class of devices defined in step 2. This means that there is no direct link between the storage pool and the tape devices that will be used. The device class definition can therefore also be seen as a barrier between the physical part of the TSM server's storage and the logical part. Include the storage pool in the TSM server storage hierarchy. A tape, or sequential, storage pool can be used for the following purposes: As a next storage pool of another storage pool (most common). As a direct destination of a clients data (less common). As a copy storage pool, used to store security copies of primary storage pools. 5. Add space to the sequential storage pool. This is done by adding volumes to a storage pool. TSM makes a difference between volumes associated to a storage pool and tape volumes located in the library. Volumes associated with a storage pool can be seen as logical volumes, pointing to physically available volumes (volumes that are in the library) by the volume label or name. You can define volumes in a sequential access storage pool or you can specify that the server dynamically acquire scratch volumes. You can also use a combination of defined and scratch volumes. What you choose depends on the amount of control you need over individual volumes.
36
Use defined volumes when you want to control precisely which volumes are used in the storage pool. Defined volumes may be useful when you want to establish a naming scheme for volumes. Use scratch volumes to enable the server to define a volume when needed and delete the volume when it becomes empty. Using scratch volumes frees you from the burden of explicitly defining all of the volumes in a storage pool. The server tracks whether a volume being used was originally a scratch volume. Scratch volumes that the server acquired for a primary storage pool are deleted from the server database when they become empty. The volumes are then available for reuse by the server or other applications. To add physical volumes to the library, they must be checked in. The check-in and check-out process will be discussed later.
Policy concepts
A data storage management environment consists of three basic types of resources: client systems, rules, and data. The client systems contain the data to be managed, and the rules specify how the management must occur; for example, in the case of backup, how many versions should be kept, where they should be stored, and so on. TSM policies define the relationships between these three resources. Figure 3-3 illustrates this policy relationship. Depending on your actual needs for managing your enterprise data, these policies can be very simple or very complex.
Policy Set
Client
Client Copy Group Policy Domain Rules Client Copy Group Client Rules Manangement Class Data Manangement Class Data
Client
TSM has certain logical entities that group and organize the storage resources and define relationships between them. Client systems, or nodes in TSM terminology, are grouped together with other nodes with common storage management requirements, into a policy domain.
37
The policy domain links the nodes to a policy set, a collection of storage management rules for different storage management activities. A policy set consists of one or more management classes. A management class contains the rule descriptions called copy groups, and links these to the data objects to be managed. A copy group is the place where all the storage management parameters, such as number of stored copies, retention period, storage media, and so on, are defined. When the data is linked to particular rules, it is said to be bound to the management class that contains those rules. Another way to look at the components that make up a policy is to consider them in the hierarchical fashion in which they are defined, that is, consider the policy domain containing the policy set, the policy set containing the management classes, and the management classes containing the copy groups and the storage management parameters.
3.1.2 Installation
This section describes the installation of the TSM server. To install the server, your system will need the following requirements: Operating system: Windows NT 4.0 SP 4 or higher Windows 2000 Server (build 2072), Advanced Server or Datacenter Server Microsoft Internet Explorer, Version 5 or higher
Software:
Components that can be installed (server part): TSM server The server code includes the TSM database and tools for server management. Installation of this part is required. TSM licenses The license package includes the code and license files required to register the licenses you purchased. If this part is not installed, the TSM server will run in evaluation mode. TSM device driver Tivoli Storage Manager provides a device driver, called ADSMSCSI. This device driver allows TSM to manage and control automated tape libraries and tape drives without the need for additional Windows device drivers. Native Windows device drivers are supported as well, but have some limitations when used for TSM, such as the fact that TSM does not know which physical device type is used, inability of TSM to detect and report I/O errors, and limited support for barcode readers. In some conditions however, it is mandatory to use the native Windows device drivers. As seen later, the LTO tape drives within the 3600 tape library series fall under this category. For an overview of supported devices and support notes, please refer to: http://www.tivoli.com/support/storage_mgr/requirements.html In addition, the TSM client code can be installed. Installation if the TSM server and related code is straightforward. When the installation CD is inserted, autorun will automatically start the Welcome window. If autorun is disabled, run the setup.exe program in the root directory of the CD. TSM will require that you define a temporary directory where the installation program will extract the installation package. After defining this directory, the extraction will take place, and the TSM Server CD Browser window will appear (see Figure 3-4).
38
To install the products, click the Install Products link. This will start the Install Products window. Here you can select the component that you wish to install. Start by selecting the TSM Server installation. After selecting the setup language (defines the language used in the installation windows) and the destination directory, you will have to choose between the Typical or Custom setup type. By selecting the custom type, you can choose the language support for the product that you require (the language in which the TSM server product will run), and eventually disable the installation of the online manuals. After completing this step, the server code will be installed. When finished, click the Finish button to go back to the Install product window. Continue the installation by starting the license installation. This step will not actually register the licenses (which will be done during the configuration step later), but only copy the required files. If this step is omitted, the TSM server will only be able to run in evaluation mode. Evaluation mode will allow you to run the TSM server for 90 days. Complete the server installation by installing the device driver. After installing the device driver, the installation program will propose a reboot of the system. If this is the last step of the installation, you can select to reboot now. If not, you can choose not to reboot the machine immediately. However, a reboot will be required before configuring the TSM server.
39
Tip: Latest version As always, it is a good idea to start the installation using the last available level of the product code. This could be done by first installing the base GA level of the TSM server as described above. A second, more convenient way however is to start the TSM server installation by simply launching the PTF (program temporary fix) installation program (IPxxxx.EXE). This will, after extracting the package, bring up an installation window similar to the GA code. The only difference is that you do not have the possibility to install the license files, which are located on the official product CD. You can install the TSM server code and the device driver, and then launch the license package installation from the GA product CD. You do not have to reboot the system after installing the device driver. Do the license installation first. Latest versions are found at: http://www.tivoli.com/support/storage_mgr/adsercli.htm This completes the installation of the TSM server. If required, you can start the installation of the TSM Backup/Archive client. As with the TSM server, you can use the available GA code from the product CD, or use the last available version. In contrast with the server, there is no need to install license files for the client. Licensing of the client will be centrally managed on the TSM server. Installation of this package is not required, and only necessary if you need one or more of the following functions: Local backup/archive client Administrative command line interface The last component will probably be chosen by more experienced TSM administrators, who might prefer using commands to configure and administer their TSM environment. If this is required, be sure to follow the Custom installation path during the client installation, since the TSM administrative command line interface will not be part of the typical installation. If you choose not to install this component, you can still administer the TSM server using the provided MMC console (see Figure 3-8) for TSM or the Web administrator.
40
Note: In order to use the TSM management console on Windows NT 4.0, the Microsoft Management Console (MMC) must be installed. TSM requires Version 1.2 or greater of the MMC. Microsoft provides a redistributable install module for MMC. The TSM package includes the redistributable install module, and the installation program will install it for you (if necessary). There are different installation packages for MMC, depending on the language version of Windows that you are running. Due to the relatively large size of the MMC packages, this package includes only the English language version of MMC. If you wish to install the TSM management console on a non-English Windows NT 4.0 system, please install the required language version separately.
3.1.3 Configuration
After installation and reboot of the system, TSM needs to be configured. This configuration consists of setting up the entire policy, licensing, clients and devices. There are two ways of configuring the TSM server: Using the configuration wizards (only available on Windows platforms) Using administrative commands Usage of the wizards has the advantage that the initial TSM setup is quick and easy. However, more advanced users might miss the granularity and control that is available when going through the configuration using commands. In this case, we will discuss the general setup using the wizards, and come back with more details when it comes to setting up the library and tape devices.
41
The following configuration steps will be necessary: Analyze drive performance to determine the best location for the disk storage pool volumes, database and recovery log. Initialize the TSM server database, recovery log and disk storage pool volumes. Register the required licenses. Configure TSM to access the storage devices. Prepare media. Register and configure TSM clients.
Initialization Wizard
The Initialization Wizard will create all required files and format all volumes required for the TSM server instance. The proposed directories where to locate these files can be changed. By default, the wizard will use c:\program files\tivoli\tsm\serverN as the instance directory (with N being the number of the instance being installed). Files in this directory will be limited in size. The location and size of the larger volumes (storagepool, database and recovery log), will be based on the result of the Performance Wizard. Again, these can be changed to the location and size chosen by the TSM administrator.
Disaster recovery manager Space management clients (UNIX) DFS/AFS clients (UNIX) There are also two licenses specific to the tape library and devices: Managed libraries Shared libraries A shared library license is required when you are sharing one or more libraries between different TSM servers (see Library Sharing on page 66). A managed library license is required on the library manager when using library sharing, but also, and more important in this context, for libraries that fall in the Extended Device Category. Table 3-1 shows an overview of the license requirements for the libraries.
Table 3-1 License requirements for TSM and IBM tape libraries Library Name 3600 Autoloader 3600 Tape library (20 slots) 3600 Tape library with one library expansion unit (LXU) (40 slots) 3600 Tape library with two library expansion units (LXU) (60 slots) 3502 Autoloader 3502 Library 4mm DDS/4 Autoloader Managed Library license required? No No Yes Yes No No No
43
There is one exception for alias names: when the Windows device driver is used to address a tape library (as shown for example in Figure 3-6), the Windows device name will be used, which has the format \\ .\Tapen, where n is the device number. Table 3-2 shows the names used for devices within TSM:
Table 3-2 Device names within TSM Name Logical device name Devices alias Windows name Format Literal (max. 30 characters) lbx.y.z.n mtx.y.z.n \\ . \Tapen Usage Used within TSM to address the library. All library related commands will use this name. Used by the TSM server to access the device through TSM device driver. Used by the TSM server to access the device through the native Windows device driver.
Always start with the library device. Next, drag the tape device on top of the library icon, which will now be located in the right pane. Figure 3-7 shows the wizard window after completing this operation.
44
Figure 3-7 Tivoli Storage Manager Device Configuration Wizard, completed view
By doing this, the wizard will perform the following operations: 1. Define a library, using the device name as TSM logical device name (in the example above, this would be LB0.0.0.1). 2. Define a device class conforming to the format used by the drive or drives located in the library. 3. Define the drive, again using the device name as logical device name. 4. Define a storage pool, using a name based on the device format or device class. More details may be found in 3.1.4, Operation on page 50.
45
As the figure shows, there are three storage pools. These consist of the two default disk storage pools (DISKPOOL and BACKUPPOOL), and a sequential tape storage pool, LTOPOOL, which is the result of running the Device Configuration Wizard. The initial hierarchy of the storage pools is as follows: The BACKUPPOOL storage pool has DISKPOOL defined as its next storage pool. This means that data residing in the BACKUPPOOL storage pool will be migrated to the DISKPOOL storage pool whenever the high-level migration threshold is reached. A sequential storage pool, LTOPOOL, is defined. Note: The wizard window will show only clients that are associated with the highlighted storage pool. In case that you have more than one storage pool, make sure you select the one which your clients are using as destination. The next step will be to modify the hierarchy in such a way that data from the DISKPOOL storage pool will also be migrated, and this to the LTOPOOL storage pool. This can be done by selecting the DISKPOOL storage pool icon, and dragging it to the LTOPOOL storage pool. The result of this will be that LTOPOOL will become the next storage pool of DISKPOOL. Figure 3-9 shows the result of this operation.
46
Figure 3-9 Client Node Configuration Wizard - setting the storage pool hierarchy
After setting the hierarchy, the next step will be to define the client policy, and add client nodes. Clicking the Add Node button will bring up the node policy information window (see Figure 3-10). You can enter the name of the node and password, as well as choose the destination for that nodes backup data.
The Policy Information tab allows you to define the policy settings for a client (see Figure 3-11). These settings include the number of versions to keep, and the period that you want to keep them.
47
Attention: When you define a new client, and modify the policy settings for that client, the wizard will create a new policy domain for that client. Although this is perfectly functional, larger installations might want to use the policy domain for a type of client, and define separate management classes for differences in data retention values. In that case, do not use the wizard to modify the retention settings, but use any of the other TSM administration interfaces.
48
When the Next button is selected, the wizard will give a list of tape volumes currently inserted in the library, and some options for the labeling process (see Figure 3-13). You can select the volumes in the provided list. Clicking the Next button will start the labeling process.
The following options are available: Overwrite existing label This options will force a label, even if the tape cartridge is already labeled. Use this option only if you are sure that there is no valid data on the tape you are using, since a label operation will erase the entire tape.
49
Barcode reader TSM provides the possibility to use a barcode reader (when available), which scans the barcode label located on the media, and uses that name as label. If this option is not used, the wizard will ask a manual input for each tape that needs to be labeled. Keep volumes in library When this option is set, the tapes will remain inside the library after being labeled. If it is not set, TSM will instruct the library to remove the cartridge after it has been labeled. Prompt user to insert volumes If set, the wizard will ask the operator to insert tape volumes before they are labeled. Restriction: During our testing, some restrictions of the Media Labeling Wizard became apparent. These include: When the barcode scan is disabled, you can label only one volume at a time. Make sure that you do not select more than one volume. The Labeling Wizard will use the device name as the library name. This is valid when the devices were configured using the Device Configuration Wizard. However, in all other situations where the TSM administrator chose to use a different logical device name for the library, the Labeling Wizard will fail. It is impossible to perform a check-in of the tapes using a private status. After labeling completes, the wizard proposes to perform a check-in operation on the volumes just labeled. This will set the status of the volumes as being inside the library, meaning that they are available for TSM to use them. The check-in process will add the tapes to the scratch tape pool. Additional wizards are available, including: MSCS Cluster Configuration Wizard (only available when the TSM server is installed on a clustered system) Scheduling Wizard (setup client operation automation)
3.1.4 Operation
This section describes in detail how to configure and use the different tape automation products discussed in this paper. They include: IBM 3600 library on page 50 IBM 3600 Autoloader on page 55 IBM 3502-x14 Library on page 62 IBM 3502-108 Autoloader on page 64 IBM DDS/4 Autoloader on page 64
50
Windows device driver The LTO drive inside the library is currently not supported by the TSM SCSI device driver. It is therefore necessary to install the native Windows device driver for the tape device. You should install the driver before starting the Device Configuration Wizard. This driver is available from IBM at the following location: http://www.storage.ibm.com/storagesmart/lto/support/lto_ftp.htm When the Device Configuration Wizard starts, Figure 3-14 is the first window that you will see. It shows the 3600 library and the tape drive as seen by the TSM server.Be sure that the device name for the library is in the TSM alias format (see Table 3-2 on page 44) and that the tape drive is in the native Windows device name format.
If the device names seen by the wizard do not conform to this naming scheme, it is an indication that there is something wrong with the device driver configuration. Start by checking that the Windows device driver is correctly installed. To do this, start the Control Panel and click the tape device icon. The IBM Ultrium Tape Device Driver should be started. If this is not the case, check that you installed the driver.
51
Another possible conflict situation can arise from the fact that the TSM SCSI device driver is trying to manage the library. Check this by starting the MMC console. Go to the service information window, and double-click the correct device driver to see the properties of the device driver service. As shown in Figure 3-15, the tape drive should be listed as controlled by native device drivers.
52
Important: Windows 2000 users In order to correctly use LTO libraries, including the 3600, under Windows 2000 in combination with TSM, it is important that you perform the following steps: Disable the Removable Storage Manager (RSM) service under Windows 2000. This is especially important when you are connecting the library through a SAN environment. Disable the Windows 2000 plug and play device configuration for the LTO tape drivers. To do this, you must edit the registry and update the device path so that it reflects the location of the device driver. The device path can be found in the following registry key:
HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion\DevicePath: REG_EXPAND_SZ:%SystemRoot%\Inf
Add the path in which the drivers are located. For example, if the drivers were located in the %Windir%\Drivers directory, the value of the key should read:
REG_EXPAND_SZ:%SystemRoot%\Inf;%Windir%\Drivers
This operation is not necessary when the device drivers for the LTO drives were added during the installation of Windows 2000. If the devices are recognized correctly, you can proceed by defining the device configuration for TSM. Start by adding the library and all the available tape devices (see Figure 3-14 on page 51, left pane) to the TSM devices (see Figure 3-17, right pane). Do this by selecting the appropriate devices and then clicking the > button.
Figure 3-17 Device Configuration Wizard, devices added to Tivoli Storage Manager
When the tape drive is added, the configuration wizard will ask you to enter the element number for that device (see Figure 3-18). An element number is used by TSM to address components within a library. These components include the tape drives, the cartridge slots and the robotics. The element number depends on the library type. Appendix A, TSM element numbers on page 101 gives an overview of the element numbers of the discussed tape libraries. In this example, the element number is 1.
Chapter 3. Software configuration
53
Once the devices are added, drag the tape devices onto the library icon in order to add them to the library (as shown in Figure 3-19).
This completes the setup of the IBM 3600 library. To use it, you must now add volumes to the library, label them, and define a storage pool that uses the defined library. For more information, please refer to 3.1.1, Concepts on page 34.
Volume handling
Typically, automated library devices require little human intervention after you set up a media rotation. However, you may occasionally need to add, remove, or manually manage media in automated library devices. TSM tracks the media in the library volume inventory, which it maintains for each automated library device.
54
You can remove volumes from automated library devices by issuing the CHECKOUT LIBVOLUME command. TSM mounts each volume and verifies its internal label before checking it out of the volume inventory. After a volume has been checked out, TSM moves the media to the entry/exit port of the device if it has one, or TSM requests that the operator remove the volume from a drive within the device. Partially written volumes that are removed from the device will need to be checked in again if TSM attempts to access them. Partially written volumes that are checked out of a library device continue to be defined to a storage pool and have a status of private. You can return partially written volumes by checking in the volume by issuing the CHECKIN LIBVOLUME command with STATUS=PRIVATE parameter. When the TSM volume inventory doesnt match the actual content of the library, errors will occur. The most common cause for these errors is manual removal or insertion of tape volumes, without going through the check-in and check-out procedure. A way to synchronize the physical content of the library and the library inventory table from the TSM server is to perform an audit of the library, using the AUDIT LIBRARY command.
Cleaning
The TSM server can control cleaning tape drives in SCSI libraries. For automated library devices, you can automate cleaning by specifying the frequency of cleaning operations and checking a cleaner cartridge into the library's volume inventory. If your library device includes its own functions for drive cleaning, you need to decide which method to use: the device's built-in drive cleaning or the TSM server's drive cleaning. Device manufacturers that include automatic cleaning recommend its use to prevent premature wear on the read/write heads of the drives. If you decide to have the TSM server control drive cleaning, disable the device's own drive cleaning function to prevent problems.
55
Figure 3-20 shows us that the autoloader device uses lb0.0.0.2 as device alias name, while the tape drive uses \\.\Tape0 as the Windows device name. This information will now be used to define the library to TSM. Start the TSM Web administrator (URL: tsm_server_address:1580) using your preferred Web browser. In the left pane, expand Object View -> Server Storage -> Libraries and Drives -> Automated Libraries. In the right pane, you will now get the Operations drop-down menu. Select Define Automated Library. This will bring up the library definition window (see Figure 3-21). Add a library name, for example LTO_Autoloader, the device name (lb0.0.0.2) and select the sharing option. Sharing a library means that the library is SAN attached and will be used by multiple TSM servers (see Library Sharing on page 66). In our example, we will not share the library. When completed, click the Finish button. This completes the library definition.
56
A second step in defining the library infrastructure is the definition of a device class. Expand the Device Classes section, and click LTO Device Class. From the Operations drop-down menu, select Define Device Class. This will open the device class definition form (see Figure 3-22).
57
Set the fields to the values you require. The device class name can be chosen by the TSM administrator. Select the correct library name from the library name drop-down menu. This is the name you gave to library when defining it earlier. All the other parameters will define how TSM will use the physical library and drives. The most important parameters are: Recording format This sets the format in which the tape drive will operate. Possible values depend on the type of device class you are defining. The default value is DRIVE, which will get the format value from the drives firmware. Other possibilities include ULTRIUM and ULTRIUMC. The last value will indicate that data will be compressed using hardware compression on the drive. Mount Limit The mount limit value indicates how many tapes can be mounted simultaneously in the library. In normal conditions, this amount is equal to the number of tape drives located in the library. The default value for this field, DRIVES, reflects this number. You can however set the value to whatever value you require. Mount Retention Indicates the number of minutes the drive should keep a tape mounted after it stops using it. The purpose of this is that the tape doesnt need to be mounted again for the following read or write operations. When a mount request for another tape is needed, TSM will dismount the idle tape, even when the retention period hasnt expired.
58
Mount Wait This parameter sets the number of minutes an operation will wait for a tape mount before generating a mount failure. Click the Finish button when completed. The next step is to define a tape drive, located in the library and using the parameters as set in the device class definition. In the Libraries and Drives directory, select the Automated Drives option. From the drop-down menu, select Define Automated Drive. This will load the drive definition form, as shown in Figure 3-23.
Start by selecting the correct library from the Library Name drop-down menu. Enter the name you want to give to the drive (for example DRIVE_IN_AUTOLOADER) in the Drive Name field, and then enter the device name, which is the Windows device name, \\.\Tape0, in the case of an LTO tape drive. The element number is 1 for this drive (see Appendix A, TSM element numbers on page 101). Complete the definition by clicking the Finish button. This completes the definition of the tape library and drive. To use it, we must now define a storage pool and integrate that storage pool in our storage hierarchy. We will start by defining a storage pool, called LTO_POOL and then update the DISKPOOL storage pool so that it uses this new defined storage pool as next storage pool.
59
The storage pool definition form for tape storage pools (sequential media) can be found by clicking Object View -> Server Storage -> Storage Pools -> Sequential Access Storage Pools. Select the Create New Storage Pool option from the Operations drop-down menu. As you can see in Figure 3-24, the definition of a storage pool has more parameters to set than the previous definitions. We will keep this example simple however, and discuss only the parameters that are required for initial use of the library. For more details on the storage pool definition and parameters, please refer to the online documentation delivered with the TSM server.
Start by selecting a name for your storage pool. As in the previous definitions, this name is chosen by the TSM administrator. Next, select the device class that will be used by this storage pool. By setting the correct device class, the TSM server will know what devices are defined in this device class. Ensure that the access status is set to READWRITE. If not, set it to that value. Two other parameters that are important are the following: Maximum Scratch Volumes Allowed This parameter sets the number of tape volumes that the storage pool is allowed to use from the scratch pool. This pool represents the tape volumes that are located in the library, and are not assigned to another storage pool. If you leave the default value of 0, you will not be able to use this scratch pool. So, set it to a value representative of the number of volumes that you will use in this storage pool.
60
Collocation The collocation parameter allows you to group data from one client on one tape. If not set, the TSM server will mix the data of all the clients on all tapes. The advantage is that the migration process will be faster, and that tape volumes will be used more efficiently. The disadvantage is that restores will take longer, since they will require more tape mounts. Click the Finish button to complete the operation. The next step in the storage pool setup is to define the tape storage pool as the next storage pool for the existing disk storage pool. To do this, click Object View -> Server Storage -> Storage Pools -> Disk Storage Pools. Select the DISKPOOL storage pool in the right hand pane. This will show the details of this storage pool. From the Operations menu, select the Update Storage Pool action. This brings up the window as shown in Figure 3-25. In the Next Storage Pool drop-down menu, select the just-defined tape storage pool (LTO_POOL). Click Finish to complete the operation.
This complete the storage pool definitions. Finally, we must now perform a label and check-in of the tape volumes that will be used in the library. In our setup, the tapes have barcode labels, which will be used as volume labels. Since there is no GUI interface available to perform the labeling, we need to use TSM administrative commands. You can enter these commands in the Server Command field in the bottom pane of the Web administrator. To start this Java applet, select the Show Command Line option from the Options scroll box (top right).
61
Enter the following command: Label Libvol LTO_AUTOLOADER Search=Yes LabelSource=Barcode Checkin=Scratch Overwrite=Yes Where: LTO_AUTOLOADER Search=Yes LabelSource=Barcode Checkin=Scratch Overwrite=Yes The name of the library Scan the entire library for eligible volumes Use the barcode printed on the tape label as volume name Add the new volumes to the scratch pool If the volume is already labeled, overwrite the exiting volume name. Use this option only when you are sure that you are labeling new volumes or volumes that are no longer needed.
The command will start the label and check-in process. When this process completes, the library is ready for operations.
62
Start by identifying the TSM device name of the library and tape drive. This can be done using the MMC console, as shown in Figure 3-20 on page 56. Another tool that exists is the TSMDLST.EXE program, located in the CONSOLE directory of the TSM installation path. This command line tool provides a text-based output as shown in Example 3-1.
Example 3-1 TSMDLST.EXE output Computer Name: TSM Device Driver: SANCLT2 Running
TSM Device Name ID LUN Bus Port TSM Device Type Device Identifier --------------------------------------------------------------------------------------lb0.1.0.4 0 1 0 4 LIBRARY ATL L500 6320000001E mt0.2.0.4 0 2 0 4 DLT QUANTUM DLT7000 2150
Once you have identified the device names, lb0.1.0.4 for the library and mt0.2.0.4 for the drive in this example, start by defining the library. This is done using the Define Library command, as follows:
define library 3502-r14 libtype=scsi device=lb0.1.0.4
The parameters given with this command include the library name we want to use, 3502-r14 and the library type, which is SCSI. There are other types of libraries, but the most common one is the SCSI type. This simply indicates that the library uses a SCSI-controlled robotics system to move the tapes. Next, define a device class. The device class definition will set the format of the tape devices located in the newly defined library (3502-r14).
define devclass DLT_class library=3502-r14 devtype=dlt format=dlt35c
The parameters include the name of the device class (DLT_class). This name can be chosen by the TSM administrator. It also contains the library to which the device class is associated (library=3502-r14), and the type of device, which is DLT (devtype=dlt). A final parameter is the format parameter, indicating the recording format that will be used (format=dlt35c). The DLT35C format indicates that we are using 35/70 GB cartridges (DLT7000), and are using compressing (indicated by the trailing C). Next, define the drives located in the library. In our example, we only have one drive. If you have more, repeat the following command, changing the parameters to the required values.
define drive 3502-r14 drive0 device=mt0.2.0.4 element=16
The command above defines a drive called drive0, located in library 3502-r14. The device name is mt0.2.0.4, as reported by the TSMDLST tool (see Example 3-1 on page 63). The element number, indicating the location of the tape drive in the tape library, is 16 for the first drive of the IBM 3502 DLT tape library. See Appendix A, TSM element numbers on page 101 for details. This completes the actual definition of the library as a storage device within TSM. To use it however, we must still define a storage pool that uses this storage device, and change the storage hierarchy in such a way that it uses this newly defined storage pool. The following command will define the tape storage pool, using the Define Stgpool command.
define stgpool DLT_Pool DLT_Class Maxscratch=10
63
The first parameter, DLT_Pool, indicates the name we want to assign to the storage pool. The second parameter, DLT_Class, defines which device class will be used by the storage pool. It is through this link that the TSM server will eventually assign the physical library and tape drive devices to the storage pool. The Maxscratch parameter indicates how many tape volumes can be used from the available pool of scratch tape volumes inside the library. Once the sequential tape storage pool is defined, we will update the existing Diskpool storage pool, setting its next storage pool to the DLT_Pool. This is done using the following command:
update stgpool Diskpool nextstg=DLT_pool
Finally, we need to add volumes to the tape library, and label them if they are new. The procedure for this is as follows: 1. Physically insert the volumes in the tape library. If possible, use barcode labels, since this will greatly simplify the label procedure. 2. Issue the following command:
label libvol 3502-r14 search=yes labelsource=barcode checkin=scratch overwrite=yes
The command above will label all volumes (search=yes) inside the tape library 3502-r14, use the barcode as label (labesource=barcode), overwrite the label if existent, and perform a checkin of the volumes, adding them to the scratch pool (scratch=yes). If you are using tapes that are already labeled, you can perform the checkin command without labeling by issuing the following command:
checkin libvol 3502-r14 search=yes checklabel=barcode status=scratch
64
Add the library and drive to the TSM devices, making sure that you drag the tape drive over the library icon to add it to the library.
This completes the definition of the library and drive. In the background, the wizard will also define a device class for the drive and library, and a storage pool that uses this device class. Next, start the Media Labeling Wizard, and follow the instructions as defined in Prepare Media Wizard on page 48.
65
Backup Sets
Backup sets are part of the rapid recovery feature. The purpose of this feature is to enable a restore of a client system without having the network overhead of sending the data from client to server. Backup sets can be seen as exported data from the TSM server, located on portable media such as tape cartridges. Once the data is exported to tape, this tape can be used locally on the client system, using the TSM backup/archive client as the interface. It allows for a restore without having to connect to the TSM server. One of the requirements is that you have a local tape drive at your client system that is able to read the tape generated on the TSM server. Currently, the following common tape media types are supported for backup sets usage: 8mm tape devices 4mm tape devices DLT tape devices 3570 and 3590 Magstar products
LAN-free
The LAN-free option allows you to back up a clients data over a SAN rather than a LAN. This requires a SAN-attached tape library, and the Tivoli Storage Manager Storage Agent product. The TSM client (including the TDP clients) will send data directly to the tape library over the SAN. The TSM server will still control the clients data, and control the library.
Library Sharing
Library sharing allows the sharing of a tape library by several TSM servers. This can greatly reduce the cost of a TSM implementation by increasing the efficiency of your library. Again, this feature is only available when the tape library is SAN attached. One TSM server will act as library manager, and will control the library. All the other TSM servers, known as library clients, can use media and drives located in the library. The following redbooks give more in-depth information about TSM: Tivoli Storage Manager 3.7.3 & 4.1 Technical Guide, SG24-6110 Tivoli Storage Manager Concepts, SG24-4877 Using Tivoli Storage Manager in a SAN Environment, SG24-6132. In addition, TSM user manuals can be viewed online, at the following location: http://www.tivoli.com/support/public/Prodman/public_manuals/storage_mgr/admanual.h tm
66
67
Other available options include: Agent for IBM ADSM Agent for Exchange Server Intelligent Disaster Recovery Agent for Lotus Domino Network Storage Executive Open File Option Agent for Oracle Server Remote Agent for NetWare Agent for SAP R/3 for Oracle Agent for SQL Server Intelligent Image Option Agent for Microsoft SharePoint Portal Server RAIDirector Note: These options (except Intelligent Disaster Recovery) are fully operational in the evaluation copy of this product. For more information, please see the VERITAS Web site.
3.2.1 Concepts
VERITAS Backup Exec lets you back up data directly to storage devices without using any intermediate stages. The current version adds support for backup to hard disks. Device and media management functions help to track usage of media. The built-in scheduler, media overwrite protection rules, and media rotation all help with the implementation of backup strategies. Backup Exec for Windows consists of the following modules: Media server The Backup Exec engine runs on the media server. It handles all jobs submitted to it, such as backup and restore, and manages all backup devices and media. Administrative console This component can run either on the media server or a remote server and is the primary interface for interacting with the Backup Exec system. It is used to submit jobs, view results, hardware utility functions, and so on. Client agents These are installed on servers and workstations and allow drives and directories on the local system to be protected by Backup Exec. Device and media management Device management functions help to organize and allocate the storage devices attached to the backup server. Backup Exec can organize drives into pools known as drive pools and cascaded drive pools. In the following sections, we will look at these components in more detail.
68
Unmodified Unmodified
Archive
Unmodified
Workstation agents
VERITAS Backup Exec provides backup agent software for backing up data from Windows, Macintosh, UNIX and NetWare systems. It is also possible to password-protect drives and directories, thus requiring the administrator to provide the password before the directory or drive can be backed up. When configuration is complete, the agents generally advertise their presence to the backup server. These workstation or server agents may be seen in the backup and restore windows, together with local and network drives.
Device management
Backup Exec automates device management and media management functions. Device management helps to organize and allocate the storage devices attached to the backup server. Backup Exec uses the concept of pools: drive pools and cascaded drive pools, with the pools containing the actual physical tape devices.
69
Drive pools These are groups of two or more drives to which jobs may be assigned and run on the first available drive. This offers fault tolerance by moving jobs from a failed drive to the next available drive in the drive pool, concurrency by running different jobs on drives in the drive pool at the same time and load balancing by distributing jobs more evenly across all drives in the drive pool. Cascaded drives These are groups of two or more drives of the same type and capacity, appearing logically as one device. Large backup jobs that exceed the capacity of the media in one drive automatically continue on media in the next drive in the pool. As such, you can create logical drives that have a larger capacity then the single drive components. The device management capabilities in Backup Exec allow you to: Monitor device usage statistics and track hardware errors Generate reports containing details of when a drive was last cleaned, the number of errors it is generating, hours of use, bytes processed, etc. Device operations Device management functions allow you to manage the physical drives and perform operations on them such as: Delete drives Pause and resume drives Inventory drives Rename drives Erase media in drives Catalog media in drives Retension tape media Eject media Label media Format media View drive statistics and properties Enable and disable drives Enable and disable hardware compression
Media management
The media management capabilities in Backup Exec help to recognize and categorize media in the storage device. Backup Exec classifies media into the following three categories: All media All media that have been introduced into the Backup Exec system. System media Media that have not been allocated to a media set, such as scratch, retired, and imported media. Reusable media are media that do not yet belong to a media set, which means new media, erased media, and media moved from another media set or group, such as imported media. Imported media are media created by another version of Backup Exec or another product. Data on imported media can be restored until the media are overwritten. Retired media are media that have been discarded because they are prone to an excessive number of errors and will not be reused. Retired media should be removed from Backup Exec.
70
User media sets This is all media that have been assigned to media sets. A media set contains the tape volumes that contain the valid data. Media sets are defined by a retention period and append period. Until the retention period expires, the media are considered allocated, after which the media become recyclable and available for reuse. Recyclable media are media whose data retention periods have expired. To protect media from accidentally overwriting media, Backup Exec uses the Media overwrite protection mechanism. This mechanism has three levels of protection. There are: Full This setting protects media that belong to an allocated media set or imported media. This is the safest option because the media cannot be overwritten unless they are moved to the scratch category, or the media have exceeded their user-assigned data retention period. In that case, the media will belong to the recyclable media category. Partial Imported and recyclable media can be overwritten. Media belonging to an active media set are protected. None This is the least safe option, and basically disables the overwrite protection feature. The administrator is responsible for making sure that media in the storage device do not accidentally get overwritten. In addition, media management provides the following functions: Catalog Erase Format Inventory Label media Rename Retension Properties
3.2.2 Installation
The installation of Backup Exec for Windows 2000 / NT is straightforward if the correct tape device drives are installed. Do not use the native drivers under Windows NT or Windows 2000, as was the case with the older versions. Since version 8.0 Backup Exec provides its own device driver files. The consequence of this is that the tape device is still listed in the Windows control panel, but no longer has a device driver associated. After you have started the Backup Exec installation process from the CD, you will first have to select the language you would like to install. After the initial information panels, the window shown in Figure 3-29 will allow you to choose between installing the Backup Exec or the Remote Administrator code.
71
The Remote Administrator provides a way to control the backup server from a remote system running Windows 2000 or Windows NT. When choosing Install Backup Exec software or options, the next installation panel will request the license information. The License panel will list the license key for any product or option previously installed. Figure 3-30 shows the selection window. In our example, we choose to install the following components: Backup Exec code Tape Device Drivers You should install the tape device drivers through this selection panel. This will perform the needed device driver installation that is used for the VERITAS software. However, some drivers will be less recent than others. It might be that you have to reinstall a newer device drive list, which can be found on the VERITAS Internet site. Online Documentation Robotic Library Support The robotic library support enables Backup Exec to work with automated library devices. As mentioned above, you will need an additional license if you have more than one tape drive installed in the library.
72
Click Next to start the installation. If you are using Windows 2000 and a device in a robotic library is discovered by Backup Exec, you will need to define if the library will be controlled by the Backup Exec or the Windows Removable Storage Manager. The choice depends on whether you are planning to share the library between several applications, or if Backup Exec will be the only one using it. If the latter is true, select the first option (which is the default).
After the installation program copies the necessary program files, you will need to register a service account that will run the required Backup Exec services (see Figure 3-32).
73
These services include: Agent browser Alert server Device and media service Job engine Naming service Server You can use an existing account or define a new one. This service account, whether existing or new, will be assigned the following rights: Log on as a service and Administrative rights. Note: If you choose an existing user, ensure that the security attributes of that account reflect that the password never expires. If you dont, Backup Exec will stop functioning when the password expires. After the account has been generated, the installation program installs the device drives for the tape and library devices. During this stage you still have the option to select what driver you would like to use. Select the VERITAS tape device drive for all tape drives (see Figure 3-33).
74
One of the main reasons for using the original VERITAS device drivers is product compatibility. The usage of those drivers ensures that the device is fully operational in the backup software. It might be necessary to uninstall existing drives or to download newer device drivers from the Internet. To find a list of the latest available drivers and patches for the Backup Exec product, please see: http://support.veritas.com/menu_ddProduct_BEWNT.htm In addition, you can click Compatibility Lists to find a list of supported devices. When the Next button is clicked, the Device Driver Installer Wizard will scan the system for a list of available devices (see Figure 3-34).
75
If the scan completes, ensure that the tape device you are planning to use is installed. If it is, click the Next button. This will install the device driver required for the detected tape device. After the drivers are installed the software will continue with installing the robotic library support and eventually the remote client agent (if required).
3.2.3 Configuration
When starting Backup Exec after the reboot, the Configuration Wizard will start. The first step in the configuration of Backup Exec is to define the way media protection will be handled (see Figure 3-35). Media protection defines which tape volumes can be overwritten.
76
Refer to Media management on page 70 for details on the protection level. Next, you will need to define the order in which Backup Exec will select tapes from the available pools. You have the choice between two types of media: recycled media or scratch media.
When the media configuration is complete, the Virus Protection Wizard is launched. This wizard allows you to set if data is scanned for viruses before taking a backup. The following wizard, the Windows Explorer Backup Wizard, allows you to add a shortcut to the Windows explorer shortcut menu, that will launch backup jobs. Next, the Device Configuration Wizard will start. The initial window will show you attached tape devices (see Figure 3-37). In this example, the IBM 120/240 GB DDS/4 Tape Autoloader is attached to the system. The DDS/4 Tape Autoloader is a so-called Multi-LUN device, which means that the tape drive and the robotics share the same SCSI ID, but use different LUNs. To use Multi-LUN devices, make sure that your SCSI host adapter and device driver support multi-LUN devices, and that the function is activated. Other libraries in the xSeries range (3600 and 3502 Series) are SCSI ID based libraries. Every component in the library has its own SCSI ID. When using separate SCSI IDs for the drives and the robot in the library, Backup Exec requires that the robotic arm has the highest SCSI ID. This means that the SCSI ID of the robot must be higher than that of the tape devices. Also, make sure that your autoloader or library is set to use Random mode. Random mode enabled the robot to be controlled by external SCSI commands.
77
Depending on the device you are using you might see different entries such as: HP 1 (Port: 1, Bus: 0, Target ID: 3, Lun 0) for the 3600 Tape library controller IBM 3 (Port: 1, Bus: 0, Target ID: 1, Lun 0) for the IBM LTO tape drive used in the 3600 Quantum 0 (Port: 2, Bus: 0, Target ID: 5, Lun 0) for the DLT tape product If the tape device is not shown in this panel, there might be a problem with the device or device driver. By pushing the Configure Devices button, the installation program will launch the Windows tape device configuration panel. Ensure that the device is seen by Windows NT, and that both device and device driver are functioning. Clicking the Next button displays the panel shown in Figure 3-38 to help verify drive configuration and ensure that drives appear correctly as drives in an autochanger or as a stand-alone drive.
78
3.2.4 Operation
When Backup Exec is launched after the initial configuration, the Backup Exec Assistant is automatically started (see Figure 3-39). This user interface provides access to all the different Backup Exec Wizards. The wizards can help and guide you through a variety of possible tasks.
79
The Backup Exec Assistant may be used for most operations. If you prefer to go immediately to the Backup Exec client interface, uncheck the Always show me this at startup button, and click Close. This will start the administrative console (see Figure 3-40). The administrative console will be the main tool to perform backup and restore operations. In addition, you can monitor running jobs and control the status of earlier ones.
If you prefer the wizard style of the Backup Exec Assistant, you can launch it by clicking the top left button. Device management (see Device management on page 69) is also called from the administrative console. This is done by selecting the Devices tab. Figure 3-41 shows the device management window. The tape library is positioned in the Robotic Library section under the server name where it is attached. At this time all tape devices are in the same drive pool, called the All Drives pool. A drive pool is used to define the destination of a backup operation. You can create additional tape pools, and drag the required devices (tape drives or libraries) into that newly created pool. If required, you can create a cascaded drive pool by dragging the tape devices onto the cascaded drive pool icon. As explained in Device management on page 69, cascaded pools only make sense when more than one tape device is available. A security feature allows you to lock or unlock a library or a tape device. This feature blocks the front panel controls of libraries and drives. This makes the library secure and will allow further access to the library only if the device has been unlocked or was restarted.
80
As you can see in Figure 3-41, the software is able to list information about media inserted in the library. Initially, the media label information is unavailable. Backup Exec will need to perform a library inventory operation, or media scan, first. To do this, right-click the slot icon, and select the Refresh operation from the pop-up menu. The result of the inventory is shown in Figure 3-42.
81
As with TSM, Backup Exec will store this inventory information in its database. To prevent inventory corruption, you must follow a certain procedure when adding or removing media. To change media in the robotic library, perform the following steps: 1. From the right pane in the Devices tab, highlight all of the slots in the view by clicking the first slot in the list, holding the shift key, and then clicking the last slot. 2. Right-click and select Export from the pop-up menu. 3. Replace the media. 4. Highlight all of the slots in the view, right-click and select Import from the pop-up menu. To update the list of media contained in the slots, perform the refresh operation. In addition, you can erase the contents of media, or perform a label operation, as shown in Figure 3-43.
This completes the setup and configuration of Backup Exec. For more information, please refer to the online manuals, available at: http://seer.support.veritas.com/docs/234125.htm
82
To install the server, your system will need the following requirements: Operating system: Windows NT 4.0 SP 4 or Higher Windows 2000 Server, Advanced Server or Datacenter Server
Note: The version currently shipped with IBM libraries is CA ARCserve 2000, which is now rebranded as CA BrightStor ARCserve 2000.
3.3.1 Installation
Before the installation of ARCserve 2000, you must first install the required licenses. To install the license files you must first run RegIT.exe, located in the root directory of the installation CD. The welcome window will appear. Click Next and you will see the available registration methods (see Figure 3-44). To install ARCserve 2000 in trial mode, select the Register Later option and click Next.
To register the product in operational mode, you will need to obtain a customer ID from Computer Associates. This can be by phone or Internet. Once you have your unique customer ID, click Register Now or Install Keycode File to register the product. The following possibilities exist: Online Express allows you to quickly register CA software using any Windows NT machine with Internet access. E-mail lets you send CA product registration request through e-mail. Remote enables you to register through a remote Windows machine. Phone provides you with the ability to call your local CA office and register your software if you do not have Internet or e-mail access. Install Keycode File installs a text file that is needed to register your software through e-mail or remote location.
83
For more information on licensing see: http://support.ca.com/ca-tlc/tlc_supp.html In the window shown in Figure 3-45, you must select the product that you wish to register. Use the drop-down menu and select ARCserve 2000 Advanced Edition, type in the registration code and click Next. If the license files are successfully created, click Next then Finish. You will have to repeat the whole process for the ARCserve 2000 Tape Library Option or any other product you wish to install and use. Tip: Since we are using automated tape solutions, you are required to install at least two products: ARCserve 2000 Advanced Edition and ARCserve 2000 Tape Library Option. You must install the Advanced Edition, becaue the Workgroup Edition will not work with the Tape Library option.
The installation for ARCserve 2000 is the same for Windows NT and for Windows 2000. If you have the software CD in the drive, it should start the ARCserve Product Explorer automatically. If not, run Pe.exe. Double-click ARCserve 2000 Advanced Edition to see the setup options as shown in Figure 3-46. Select the required setup option (in this case we chose Standard Setup) and click Install.
84
The rest of the installation is straightforward and similar to most other software installations. Accept the terms of the license agreement, fill in customer information and select the destination folder or accept the default. Next, you will have to select what database you will use as the ARCserve database. The ARCserve database records job information for each job that is run, including media and media device information. The database information allows you to perform restores by keeping track of each file and directory that was backed up to media. When you want to restore a specific file, the database determines which media a file is stored on. The database information is also used to generate ARCserves reports. As shown in Figure 3-47, you have the choice between the native ARCserve database or a Microsoft SQL database.
85
The next step is to specify a system account. This account will be granted administrator and backup operator rights and must be an existing system account. The account will be used to run the ARCserve services. Click Finish to complete the installation of the advanced edition component. Restart the system. Next, install the ARCserve 2000 Tape Library Option. Since the installation of this component simply copies the required files to the system, no further details are necessary. After the installation, you might get prompted for the device configuration. If this happens, exit and reboot the system.
3.3.2 Configuration
After restarting the system, you can start configuring the library and tape devices. You have several options for doing this. These include the ARCserve Device Configuration application or Device Wizard. The ARCserve Device Configuration and the Device Wizard are both available from the Start menu. Note: Unlike TSM and Backup Exec, ARCserve requires Windows Removable Storage Manager as library manager. After the initial welcome window, you will need to select the type of device that you are installing (see Figure 3-48). Select the Tape and Optical Library to configure a tape library.
86
The next window will allow you to assign drives to the library devices. This allows ARCserve to know which drives are available inside a library. The left pane should display the library devices and the right pane displays the available tape devices (see Figure 3-49).
87
You can add tape drives to the library using a manual selection and assignment, or choose the automatic Auto Config option. When performing the operation manually, select the drive and library, and click the Assign button. If you have more that one tape device in the library, make sure that you add them in the order that they are assigned in the library. For example, if your library has a Tape0 and Tape1 device, add Tape0 first. To determine which tape is which device in the library, use the SCSI ID available in the Detail window. Note: If your tape engine is started, a window appears that will let you stop the engine. The engine must be stopped in order to configure a library device. The result of the operation is shown in Figure 3-50. The tape drive should now appear under the library in the left pane of the window. Once you have assigned the devices, click Next.
Step 2 allows you to assign a cleaning tape and add barcode support. Select the library device and click Properties. Finally, the advanced configuration window appears (see Figure 3-51).
88
This window provides the following advanced options: Library Quick Initialization This allows ARCserve to skip the library inventory process that is automatically activated when the tape engine is started. This feature assumes that no media have been moved, added, removed, or swapped since the last shutdown. Mark Drive(s) Online/Offline This allows you to mark defective drives offline. ARCserve will continue to operate using the remaining non-defective drives. To finalize the device configuration, click Next. This will show you the summary of your operations. Click Finish to complete.
3.3.3 Operation
If you open the ARCserve Manager from the start menu, this will give you access to all of ARCserves functions via the menu on the left-hand side under Contents.
89
ARCserve management functions can be accessed in several ways, either by Quick Access or using the wizards. Both can be accessed from the home page or from the links in the left pane. We have used the Quick Access method as shown in Figure 3-52. In this window you can view and access all the Managers below: Backup Manager The Backup Manager allows you to schedule and submit jobs to back up your data to media. Restore Manager Allows you to restore previously backed up data. Database Manager Displays information from the ARCserve database. Job Status Allows you to monitor, submit, start, stop, reschedule and delete jobs. Devices Manager As well as displaying information about your storage devices and media, the Device Manager also allows you to perform media functions such as compression, formatting, erasing, ejecting and retensioning. Report Manager Uses data in the ARCserve database to generate reports.
90
Media Pool Manages, creates and maintains logical groupings of media for easy identification of backups. ARCserve Server Admin Allows you to manage the three ARCserve Engines or Services: Job, Tape and Database. Alert Manager Sends messages/alerts about specific events that occur during ARCserve's operation.
91
92
Chapter 4.
Recovery
This chapter discusses various types of recovery that might be necessary after data loss. It explains the different types of recovery, including business recovery, system recovery, application recovery and data recovery. For each type, we provide you with an overview of the options you have, and the necessary components. We do not go into detail on the different recovery processes or products, mainly due to their extensive nature.
4.1 Introduction
One of the major goals of implementing a backup solution is the ability to recover from disasters, hardware failures, system failures and user errors. Where tape libraries and associated software are certainly a component in recovery operations, there are a lot more components that create a valid and working recovery. These components depend largely on what type of recovery you are trying to implement. We can distinguish four types of recovery, each with its own specific needs and considerations: Business recovery System recovery Application recovery Data recovery Recover an entire business after a major disaster. Recover one or more IT systems after system failures. Recover an application after data loss or failure. Recover data after data loss (accidental or failure).
As you can see from this list, the scope of what we are trying to recover differs widely. In the following sections, we take a brief look at these types of recovery. One constant in recovery is planning. A recovery plan should clearly define procedures that need to be taken in the event of a certain condition. The most advanced hardware and software will not serve its purpose if nobody has a clear idea on how to recover. Recovery is typically a high-stress situation, and having a procedure ready makes one less thing to worry about: how will I do it?
93
A second very important factor is that even the best recovery plan is not worth a lot when the procedures haven't been tested and the plan is not maintained. Testing a recovery plan ensures that the decisions that will be made in case of recovery will be the correct ones. And, since current IT environments are very dynamic, maintaining that plan and testing the updates will ensure that the recovery procedures are up to date. Another advantage of regular recovery testing is that the people involved will not face unknown situations in the event of a disaster or failure. Having run through the tests, they will be more familiar with the procedures.
94
These problems lead backup software manufacturers to provide tools or add-on products that allow system recovery.
Chapter 4. Recovery
95
BMR Server
Boot Server
File Server
TSM Server
OS, BMR & TSM code needed at restore time Boot Image All Backed-up client files
The BMR main server manages the clients supported by the BMR system, the process of preparing for the client restoration, and the post-processing after the client(s) have been restored. It makes the appropriate boot image and file systems available to the client, ensures that the boot server and file server are properly configured, and generates a customized client recovery procedure. The boot server provides the boot image to the client when it performs a network boot prior to restoration. BMR uses standard protocols to network-boot the client from the boot server. The file server provides the client with the programs, libraries, and configuration data it needs to perform the recovery procedure. These include the necessary operating system commands and libraries, the BMR client package and the TSM client package. The BMR client is any machine being protected by BMR. The client continues to back up to its normal TSM server. In order to be able to recover the machine using BMR, all of the clients files must be backed up on a regular basis.
96
97
98
files. The advantages of this system over the real offline backup are that the required offline time for the database application is very limited. A disadvantage of this backup method is that it creates a two-phase backup and restore mechanism: first dump the database, then backup it. Or in case of a restore: restore the dump file to disk, and then import it into the database. The online backup method produces the best result when working in an environment where 24X7 availability is required. The backup program will allow to produce a backup while the database is online. Another advantage is the close integration of the backup product and the database application. In some case, this integration even allows the DBA to use the database manager interface to perform backups, restores and check the status of the previous backups (an example of this is the Tivoli Data Protection for SAP/R3 product). A disadvantage compared to the offline backup solution is that you will need additional software in most cases, making the solution more complex and expensive. Most backup software manufacturers provide online database backup tools. Tivoli for example has an entire range of Tivoli Data Protection for Applications products, integrating online application server backups with the Tivoli Storage Manager product. A special application where you would certainly need a recovery scenario is the backup server itself. It is logic that when this component fails, all restores become impossible. That is why most constructors provide clear and easy recovery steps for the backup server. The important part is to check if your installation complies with the prerequisites.
Chapter 4. Recovery
99
100
Appendix A.
101
Drive 1 (1)
Drive 2 (2)
40
41
Robot (0)
33 32 31
48 49 20 (EE)
Notes: 1. Barcode supported 2. When used with 3600-LXU expansion unit(s), the following element addresses apply: Drive 4: Element number 3 Drive 5: Element number 4 Drive 6: Element number 5 Drive 7: Element number 6 Slot element addresses for expansion unit 1 : 51 - 69 Slot element addresses for expansion unit 2 : 71 - 89 3. Entry/Exit (EE) slot is always 20, even with 1 or 2 LXUs attached, and will be the top right slot.
102
Robot (0) 36 Slot 6 35 Slot 5 34 Slot 4 33 Slot 3 32 Slot 2 31 Slot 1 Drive 1 (1) 37 Slot 7 38 Slot 8 39 Slot 9
Notes: The following options should be set during configuration: Enable Barcode Autoload Cartridge No Auto Cleaning No Random Mode
103
Notes: 1. Random mode should be selected to run Tivoli Storage Manager. This is the default. Autoload cartridge should be set to NO on the operator's panel. 2. Barcode should be enabled on the operator's panel. Ensure that the autoloader door is closed. 3. Auto cleaning is supported. Set Auto Cleaning on the operator's panel and ensure that there is a cleaning cartridge in physical slot 8 (Next to the drive). 4. Autoloader firmware level should be at 0000F, and the DLT7000 drive level should be at 1E48.
The tape drive and the autochanger share a SCSI ID, but have different LUNs. Usually the drive is LUN 0 and the autochanger is LUN 1.
104
Related publications
The publications listed in this section are considered particularly suitable for a more detailed discussion of the topics covered in this Redpaper.
IBM Redbooks
For information on ordering these publications, see How to get IBM Redbooks on page 106. Netfinity Tape Solutions,SG24-5218 The IBM LTO Ultrium Tape Libraries Guide, SG24-5946 IBM Tape Solutions for Storage Area Networks and FICON, SG24-5474 Implementing IBM LTO Tape in Linux and Windows, SG24-6268 Using IBM LTO Ultrium with Open Systems, SG24-6502 IBM Magstar Tape Products Family: A Practical Guide, SG24-4632-03 Tivoli Storage Manager 3.7.3 & 4.1 Technical Guide, SG24-6110 Tivoli Storage Manager Concepts, SG24-4877 Using Tivoli Storage Manager in a SAN Environment, SG24-6132
Other resources
These publications are also relevant as further information sources: LTO Autoloader Users Guide LTO Library Users Guide IBM 3600 Fibre Channel Card Users Guide IBM 280/560 GB DLT Autoloader Users Guide IBM 490/980 GB DLT Library Users Guide IBM 120/240 GB DDS/4 Autoloader Users Guide These publictions are found at the following Web site: http://www.ibm.com/support/
105
You can also download additional materials (code samples or diskette/CD-ROM images) from that site.
106
Special notices
References in this publication to IBM products, programs or services do not imply that IBM intends to make these available in all countries in which IBM operates. Any reference to an IBM product, program, or service is not intended to state or imply that only IBM's product, program, or service may be used. Any functionally equivalent program that does not infringe any of IBM's intellectual property rights may be used instead of the IBM product, program or service. Information in this book was developed in conjunction with use of the equipment specified, and is limited in application to those specific hardware and software products and levels. IBM may have patents or pending patent applications covering subject matter in this document. The furnishing of this document does not give you any license to these patents. You can send license inquiries, in writing, to the IBM Director of Licensing, IBM Corporation, North Castle Drive, Armonk, NY 10504-1785. Licensees of this program who wish to have information about it for the purpose of enabling: (i) the exchange of information between independently created programs and other programs (including this one) and (ii) the mutual use of the information which has been exchanged, should contact IBM Corporation, Dept. 600A, Mail Drop 1329, Somers, NY 10589 USA. Such information may be available, subject to appropriate terms and conditions, including in some cases, payment of a fee. The information contained in this document has not been submitted to any formal IBM test and is distributed AS IS. The use of this information or the implementation of any of these techniques is a customer responsibility and depends on the customer's ability to evaluate and integrate them into the customer's operational environment. While each item may have been reviewed by IBM for accuracy in a specific situation, there is no guarantee that the same or similar results will be obtained elsewhere. Customers attempting to adapt these techniques to their own environments do so at their own risk. Any pointers in this publication to external Web sites are provided for convenience only and do not in any manner serve as an endorsement of these Web sites. The following terms are trademarks of other companies: C-bus is a trademark of Corollary, Inc. in the United States and/or other countries. Java and all Java-based trademarks and logos are trademarks or registered trademarks of Sun Microsystems, Inc. in the United States and/or other countries. Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the United States and/or other countries. PC Direct is a trademark of Ziff Communications Company in the United States and/or other countries and is used by IBM Corporation under license. ActionMedia, LANDesk, MMX, Pentium and ProShare are trademarks of Intel Corporation in the United States and/or other countries. UNIX is a registered trademark in the United States and other countries licensed exclusively
107
through The Open Group. SET, SET Secure Electronic Transaction, and the SET Logo are trademarks owned by SET Secure Electronic Transaction LLC. Other company, product, and service names may be trademarks or service marks of others.
108