TSM Symposium 2007 : Preparing the Path

Proceedings from the symposium ...

Speaker Agenda Topic Topic Abstract
Michael Abel & Bruno Friess TSM Mysteries Some values of the TSM environment are very well documented and there are several recommendations from the lab. But do you really know how they influence your daily TSM work? We will talk about some famous parameters such as BUFPOOLSIZE and TCPWINDOWSIZE, including how to measure values, show the impact of a badly chosen values, and prove values work as expected, along with hints on "Best Practice".
David Asbury Linux User Experience At CERN, TSM servers have been run for some years on AIX machines. More recently TSM servers based on Redhat Linux have been added following CERN's extensive investment in Linux-based computing. The hope is this will be a cost effective way of handling the large increase in anticipated backup and archive data in future. This session will include the use of fibre-channel with RAID arrays for staging and TSM databases and how the various machines share these facilities. Our experience with Linux-based TSM servers will be presented in this context.
Gerd Becker TSM 5.4/5.5 Experiences This session will include real-life experiences of some of the new function in TSM 5.4/5.5. It will focus on active data pools and backupset enhancements (including some of the pitfalls) giving hints on how to exploit the new features.
Stefan Bender Facts and News about TSM HSM: Where are we and where will we go? The presentation will give an overview on the value of HSM for a customer. It will compare the two existing HSM products (Space Management for UNIX and HSM for Windows), explain their different approaches and how we plan to bring both products closer together in 5.5. The presentation will also mention candidates for the next release of both products. This presentation will also cover the GPFS and HSM ILM to tape feature.
Dave Canan TSM Performance This session will review disk configuration performance issues and considerations when using Tivoli Storage Manager and various disk subsystems, primarily focusing on the DS4800 disk subsystem. Topics include TSM I/O characteristics, hardware considerations, TSM disk tuning with AIX, use of SATA and SVC with TSM, and ways to monitor disk I/O performance.
Dave Cannon Data Deduplication Data deduplication offers the possibility of storing more data on disk, thus increasing the likelihood that data can be restored from disk rather than tape. This session examines technologies and design considerations for data deduplication. The expected design and behavior for TSM's future deduplication function are also described.
Dave Cannon Preview of Future Enhancements This session previews potential future enhancements in Tivoli Storage Manager. Topics include upgrade of the TSM database engine, management of data in server storage, improved administration, and enhanced security options.
Dave Cannon Understanding Disk Storage With the trend toward expanded usage of disk for backup/recovery, understanding the nuances of TSM disk storage becomes increasingly important. This session contrasts design features and trade-offs for random-access and sequential-access disk in TSM. Recent and planned product changes related to disk storage are also described.
Christina Coutts TSM Encryption Options using IBM tape encryption capability In the light of increasingly stringent legislative requirements for data security, encryption has become a critical topic. This session provides a view of the IBM tape hardware encryption offerings, and their implications for TSM installations needing to secure tape data using encryption. It includes an overview of the different methods open to TSM users, but covers in more detail the TSM exploitation of the encryption hardware now provided in the IBM tape drives.
Alexander Dunaevskiy Three Generation of Linux Servers as TSM-Platform at LRZ LRZ runs a big facility for backup and archive of data from more than 4500 nodes, storing 2 PB and 2.5 billion files. In 2003 we asked ourselves, is Linux ready for use in such big environment? We ran AIX and Linux in parallel initially and studied the differences in behaviour of TSM Server. In this session we will discuss historical and technical aspects of the use of Linux as base operational system for TSM Server. Also covered is the backup strategy of a Netapp Filer farm also on a Linux based TSM server with shared library and the experiences we had with it at LRZ.
Thomas Eifert Experiences with backup in the petabyte range Since recently RWTH Aachen University is doing parts of the backup service for other universities of North Rhine Westphalia across the wide area network. Several degrees of service are implemented, from TSM-server-to-server-backup as a fallback for local services, up to the full service to the local clients. In this presentation, the various configuration details and experiences made so far will be described.
Stephen Firmes Benefits of Virtual Tape Libraries and data de-duplication technologies with TSM In this presentation, the benefits of virtual tape libraries and data de-duplication technologies with TSM will be described. It addresses the business and resource implications of virtual tape library systems in TSM environments. This is a highly technical presentation including discussion of topics ranging from hints on reducing backup windows to improving reclamation processing.
Christian de Groot Setting up TSM Journal Based Backups on Microsoft Clusters TSM offers journaling capabilities for AIX and Windows clients having millions of files. The Journal Based Backup feature can be implemented quite easily using the built-in Setup Wizard or the dsmcutil command line interface. Setting up a high availability environment is challenging. In this session, how to setup multiple JBB services to monitor each cluster disk resource with the ability to fail over each separately is covered (using some hard-to-find information and several attempts)!
Del Hoobler TSM for Copy Services and TSM for Advanced Copy Services - Today and Tomorrow This presentation covers details on how TSM for Copy Services and TSM for Advanced Copy Services works utilizing snapshot technology for near-zero impact backup and Instant Restore operations. It takes a deep dive into TSM for Copy Services, most specifically with Data Protection for Exchange. It also discusses what is available today and what the future might hold.
Del Hoobler TSM for Mail - Data Protection for Exchange - A Look Under the Hood This presentation covers details on how to use Data Protection for Exchange to protect your Microsoft Exchange Server environment. It focuses on traditional "legacy"-style operations and lightly touches upon the new snapshot (VSS) support as well as provides some comments and discussion about "IMR".
Dorin-Daniel Lobontu TSM and dCache in grid projects at Forschungszentrum Karlsruhe The Forschungszentrum Karlsruhe in Germany has chosen TSM, for its Gridka project, to archive several PetaBytes of data. The working storage agents and TSM server use solely Linux as the operating system. The presentation focuses on: TSM layout for a large amounts of data; storage agents and lanfree data streams; SANdiscovery feature of TSM and its importance in a large and complex environment; TSM configuration in an installation with low number of drives and a sustained stream of data; TSM migration from Linux 32 bit to Linux 64 bit.
Peter McNamara The Benefits of Virtual Tape in a TSM Environment (including a Tape Technology Update) At the time of the last Symposium attendees had the choice of the IBM 3592-J1A Tape Drives with 300 GB per cartridge, or LTO3 Drives at 400GB per cartridge. Now there is the 3592 Gen 2 drive (the TS1120) at 700GB per cartridge and LTO4 at 800GB, both of which can work with a Virtual Tape Library. So, which drive is right for you, and what role can Virtual Tape play? Peter's presentation explores the options.
Stephan Peinkofer Exploiting the full power of modern industry standard Linux-Systems with TSM The main goals of this linux-focused presentation are: Presenting the core parameters which affect TSM backup and archive performance the most; Showing benchmarking tools to measure the impact of certain settings for these parameters; Proving the effectiveness of the settings through real world examples from our TSM-Server farm, currently consisting of approximately 34 TSM-Servers providing backup and archive services for over 4500 clients which generate roughly 10 TB of data per day.
Peter Pijpelink Any 2 Any Implementing TSM export with high-availability and low-bandwidth (without using a SAN)
Peter Pijpelink VTL Performance/comparison In the first part of this session VTL will be compared with other techniques, and the results of real live testing with Falconstor based VTL and Gresham VTL will be presented. The differences between random access disk, filedev and VTL techniques will be discussed. Details such as file size and load on system will be shown.
In the second part, the customer setup is described where money for remote sites was an issue. A solution was made where the main site makes all client backups over the internet and a system setup to export node data server-to-server and using the remote site as DR without using SAN.
Klemens Poschke System Storage Archive Manager - Basics and lessons learned with ISVs implementing the TSM-API This presentation covers basics on the IBM Archive Backend Storage solutions based on SSAM and DR550 and points out some of the experiences we faced during recent implementations with ECM solutions from IBM and other ISVs. It will also cover application migration scenarios and attachment of non-IBM archive appliances like EMC Centera and NetApp Snaplock.
Jarrett Potts CDP Jarrett will present a brief overview/update on CDP, give a demo and lead a discussion session.
Jarrett Potts (presented by Peter Pijpelink) Tivoli Storage Manager for Microsoft SharePoinT Come learn about the newest addition to the TSM family - TSM for Microsoft SharePoinT. This session will cover the basics of the Sharepoint Application, how TSM protects it, what is the function TSM for SharePoint provides and what are the benefits of using TSM for Sharepoint vursus the built-in produt that cames with sharepoint. We will also discuss how it is installed and configured as well as how the application is recovered.
Andy Raibeck A little SQL now and then is relished by the wisest admin One of the most frequently asked class of questions about Tivoli Storage Manager is how to query the TSM server database in order to generate custom reports. This session will focus on TSM's built-in support for the SQL SELECT statement: SELECT tutorial; Sample SELECT statements; Q & A.
Andy Raibeck Ode to ODBC (or, "OK, I installed the ODBC driver... now what?") One of the most frequently asked class of questions about Tivoli Storage Manager is how to query the TSM server database in order to generate custom reports. The answers often entail some form of the SQL SELECT statement, usually issued from the TSM Administrative command line interface. Despite having an ODBC driver since Version 2, it seems that the ODBC driver is among the least understood of TSM's many standard components.This session will focus on the TSM ODBC driver. Topics will include: What ODBC is and how it works; How to install and configure the TSM ODBC driver; Sample usage with C, Java, Perl, VB Script, and Microsoft Excel.
Andy Raibeck Potpourri du Backup The Tivoli Storage Manager backup-archive client includes a wide variety of options and techniques for backing up your data, including full incremental (progressive) backup, incremental-by-date backup, adaptive subfile backup, journal-based backup, and others. This session will discuss and compare the various backup techniques and related options. Symposium delegates will come away with a better understanding of how to approach their backups.
Freddy Saldana The Requirements Process The requirements process - including requirements gathering, prioritization, and scheduling.
Freddy Saldana Trends and Directions This session will provide an overview of the candidates for TSM enhancements, with the focus on the content of TSM 5.5 and the next release. TSM 5.5 will include changes in the areas of sequential disk pools, export/import usability, VMware Consolidated Backup, AIX WPAR and VSS support for SQL and Exchange 2007. A summary of the next release candidates will also be discussed.
Gerhard Schneider Reflections Reflecting on TSM and its place as part of our cultural heritage.
Charles Silvan Backing up very large databases This session is based on a real customer case and describes the setup to required for a production environment to backup and restore very large SAP DB2 Databases (10's of TB's). The environment, configuration, performance, optimization, disaster recovery, results and experiences will be covered.
Fiona Speakes Virtual Tape Libraries: Improving Backup Productivity in a Multi-Data Centre Environment What is a Virtual Tape Library (VTL)? Why are we looking at VTLs? Why are VTLs of interest? This session will discuss the benefits of VTLs that apply to our organisation, including: Using VTLs for offsite backups, reducing backup times and network usage between data centres; Long-term storage of archived data in comparison with traditional usage of "real" tape; Usage as short-term immediate onsite storage, with de-staging to "real" tape as an onsite storage pool.
Denise Tacay Self-care portal for Backup administrators This session describes a self-care portal which has been developed at RWTH and can be accessed via internet. The portal offers an overview of the associated nodes to the backup administrators and allows them to perform certain changes like adding or deleting a node and resetting the password. The TSM server cannot be accessed directly through this interface, but changes made via the portal will made on the server.
Josef Weingand The right Storage and the right Architecture for your Backup/Restore Environment The advantages and disadvantages of disk and tape in the backup/restore area will be discussed, how to calculate the total cost for ownership for disk and tape, where backup to disk or where backup to tape makes sense and the combination of disk and tape (D2D2T) with TSM. Also covered is the usage, advantage, and disadvantage of VTL in a TSM environment.
Jos van Wezel TSS, an interface to TSM for storage-management middleware The Large Hadron Collider at CERN will produce several PetaBytes of data each year and be held at sites such as KIT. The data management middleware developed for the experiment manages disk storage and relies on tape based systems, either to offload data or for archival purposes. At KIT the mass storage is handled by TSM which has been coupled via a software layer with the TSM API to the storage management middleware. This presentation describes TSS: an interface to TSM which uses the TSM API to read and write from tape storage and queues migrate and recall requests. Discussed are the use and benefits of the API and the purpose and effect of the TSS layer. Ideas for improvement and future development plans conclude the talk.


Back to top