2002 The Year of Archiving by bsw17644

VIEWS: 0 PAGES: 48

									2002: The Year of Archiving

      Robert Parkinson
        PCS Limited
  “Creating Solutions Through Knowledge”



                                           1
What We’ll Cover

   Archiving benefits
   Definitions of terms
   Archiving features
   Archiving steps
   Storage and management
   Archiving by module
   What next?
                             2
    What is Archiving?

 Move data no longer required in database to archive files
 R/3 Database        Archiving Program
                                              Archive Files




                      Archiving Session


                                                              3
    Why Archive?
   Volume of data in the database increases
       Disk space
       Performance
           Size of index trees, full table scans, cache hit ratios, ...
       Effort and resources required for database
        administration
           Backup/restore, reorganization, create index, statistics, ...
   Business and legal requirements
       Auditing
       Record-keeping requirements
       Potential re-useability of data
                                                                            4
 Data Continues to Grow
             Yearly data growth 400 GB, retention period 6 months
             Data not needed online after 5 years: 1,800 GB
             Volume without data archiving in 5 years: 1,000,000 GB
2000
             Volume with data archiving in 5 years:      225,000 GB
                                                                  Not needed
1750                                                              online
1500                                                              Needed online
1250
1000
                                  No backup               Backup, restore and
750
                                  necessary with          database administration
500                               data archiving          time grows proportional
                                                          to database size without
250                                                       data archiving
  0
       1996        1997    1998    1999     2000   2001

                                                                                5
    Archiving Requirements
   Independence
      of database version

      of version and customizing of application software

      of hardware (codepage / numeric format)

   Auditability
      Revenue authorities – IRS, auditors

      Data security officials

   Integration in applications
      Direct access

      Reporting functions

   Performance and robustness
                                                            6
What We’ll Cover

   Archiving benefits
   Definition of terms
   Archiving features
   Archiving steps
   Storage and management
   Archiving by module
   What next?
                             7
         Optical Archiving


   ArchiveLink can be used to archive
    documents
   Original documents or print lists can be
    stored in optical archive systems and
    displayed, but cannot be reloaded or
    analyzed
   Archive files which were created by R/3
    data archiving can be stored, read, and
    reloaded using ArchiveLink


                                               8
Database Reorganization

   Should be performed after archive
   Compelling reasons:
       Tables keep free space allocated after
        archive
       Gives free space back to the database
       Improves performance
       Reduces database maintenance times

                                                 9
    Retention vs. Residence Period

   The retention period is the entire time that
    data spends in the database before its
    archived
   The residence period is the minimum length
    of time that data must spend in the database
    before its eligible to be archived




                                                   10
    Backup & Restore

   Contains a copy of the database contents for
    security
   Backups are usually made at regular intervals,
    according to a standard procedure
   Reloading the backup into the database is referred
    to as restore
   In earlier SAP releases, the archiving procedure
    required a database backup. SAP states this is no
    longer necessary in Release 3.X and 4.X


                                                         11
What We’ll Cover

   Archiving benefits
   Definition of terms
   Archiving features
   Archiving steps
   Storage and management
   Archiving by module
   What next?
                             12
    Archiving Features

   High data security during archiving
   Optical archive connection via ArchiveLink
   Data compression
   Online data archiving
   Archiving without database backup
   Random and sequential access to archived data
    objects, reload to the database possible
   Automatic conversion of old archive files
   The ADK (Archive Developer’s Toolkit)

                                                    13
     Security Provided in Archiving

   Archiving is a two step process:
   Create archive files
   Read data from file and execute delete program
   Storage of the archive files on other media
   Provides data security in case of problems during
    the archiving process


                                                        14
Step 1 Create the Archive File

                        Archiving
                        program




 R/3
Data-                 Archive
base                    file




                                    15
        Step 2 Execute Delete Program

                                                     Archiving
               R/3                                   program
              Data-
              base

 R/3
Data-
base                    Delete                     Archive
                        program                      file
                                                          Archive
                                                            file

           Delete                 Archive
           program                  file
              Delete                     Archive
              program                      file

                                                                 16
Step 3 Store the Archive Files

                  •ArchiveLink
                  •HSM-Systems
 R/3
Data-
base




                             Archive
                               file
        Delete                      Archive
        program                       file



                                              17
           Optical Archive Link

 R/3 System         ArchiveLink
                      The archive files can be stored on tertiary
                       storage media
                      Files can be stored manually or
                       automatically
                    Hierarchical storage management systems
                       (HSM systems)
ArchiveLink            Archive files can also be passed to an
                        HSM system
                       HSM systems can use optical and
                        optomagnetic storage media
                       The HSM system manages the archive
                        files automatically
 Archive
 system
                                                                 18
Compression

Data is automatically compressed (Up to a
  factor of 5 or greater)


Online data archiving
   Data archiving may run in parallel to dialog and batch processing


Data archiving without database backup
   Data archiving is fully recoverable

                                                                        19
    Access to Archived Data

   There are two ways to access archived data:
        Sequential access to a dataset (reporting)
        Random access to individual data objects
   Archive files can be
        Read
        Analyzed (for reporting purposes)
        Reloaded (Not Recommended – only right after archiving!)
   Actions depend on archiving object
   Access for reporting and reloading requires that the file is in the
    file system
   Access to individual objects depends on an index
                                                                          20
     Conversion of Old Archive Files

   Necessary conversions are
    performed automatically
   Temporary conversion during
    data access
   In rare cases, conversion
    programs are provided




                                  21
The ADK
R/3 System           Databas
                     e
  L
                                                Archiving
  D            ABAP/4 Program
                                                Class
  B


          Codepage, Number Format, Structure Conversion
  ADK:    Direct Access, Compression, Batch-Handling, . . . .




 ArchiveLink          Archive File

                     HSM             manually

        Archive System
  with Tertiary Storage Media
                                                                22
    What is an Archiving Object?
                                      Archive

   Key to archiving concept
   Description of database objects
   Collection of programs


     Database




                      R/3 System


                                                23
     Customization During Data
     Archiving
   Creating an archiving object
        Define archive object using transaction AOBJ
        Create programs for archiving
              Archive program
              Delete program
              Reload program
              Index program
        Create programs for accessing archived data
              Reports
              Direct access via online transaction

                                                        24
What We’ll Cover

   Archiving benefits
   Definition of terms
   Archiving features
   Archiving steps
   Storage and management
   Archiving by module
   What next?
                             25
    Archiving Personnel

   Archiving can be performed
      by the Basis Administrator
      by the application department

   Motivation (and point of entry) for both are different
   Basis Administrator
      The Basis Administrator checks the database/table size
      Archiving object relationship has to be determined

   Application Dept.
        Application Dept. performs archiving for various reasons
         (periodically – month end, quarter end, year end)
        Application Dept. uses direct point of entry

                                                                    26
  Archiving Flowchart
                                         Archiving procedures


Before 1st use of archiving   Before 1st use of particular archiving object      For each session


    Coordinate activities                  Coordinate activities                 Coordinate activities

    Check dependencies                    Check dependencies                   Check space on harddisk

                                      Check application customizing           Schedule archiving session

                                         Check ADK customizing                 Check archiving results

                                                                                  Save archive files
                                            Platform independent filename

                                                    Archive file size

                                                      ArchiveLink

                                                    Delete program




                                                                                                           27
    Identifying Critical Tables

   Database Performance Screen (DB02)
       Overall database growth analysis
       Determine size of tablespaces and tables
       Free storage space – allocated vs used space
   Table Call Statistics Screen (ST10)
       Locate the most active tables
       Which tables are growing
       Which tables have slow response


                                                       28
    Identifying Critical Archiving
    Objects
   Use archive administration (SARA)
       Browse through the archiving objects
       Determine which tables are in the archiving objects
       Schedule all required archive programs
   Use CCMS database administration (DB15)
       Determine which archiving objects use a table
       Check online space information for all tables of an
        archiving object


                                                              29
    Business Requirements for
    Archiving Objects (SD_VBAK)

   Document has status "Complete"
   All documents in document flow have
    status "Complete"
   Financial document is balanced
   Additional customized checks possible



                                            30
What We’ll Cover

   Archiving benefits
   Definition of terms
   Archiving features
   Archiving steps
   Storage and management
   Archiving by module
   What next?
                             31
    Storage & Management
   Automated storage
       Low management effort
   Manual storage
       Data medium maintenance
       Data medium management
       Archive production



                                  32
    Storage of Archive Files

   Choice of medium is critical for:
      Utilization of the functionality offered by SAP

      Effort required for maintenance

      Effort required to access data

      Durability of archive

   Protection of investment by means of combined methodology
      Data archive files

      Workflow for incoming and outgoing records

      DMS (Document Mgmt. System) technical documents

      Printed reports

      Automation of database backup and restore



                                                                33
    Archive Files on Hard Disk

   Advantages
       Single document access < 5 seconds
       Quick access for reporting
   Disadvantages
       File system limitations
       Considerable manual management effort required
           Replicate the archive files
           Storage of archive files
           Data security


                                                         34
    Archive Files on HSM System

   Advantages
       Single document access < 30 seconds (with block
        access)
       Quick access for reporting (with block access)
       Scalable file system
       Long data medium retention period
       Data medium management support
   Disadvantages
       Can be a costly solution

                                                          35
    HSM System Requirements

   "Write through" should be possible
       Increased data security, as the archive files
        migrate before the delete program runs
       Data retained in "magnetic cache" as copy
   "Random access" to files must be possible
   "Hardware sharing" is useful
       Use of a jukebox via both HSM and ArchiveLink
   "Media management" should be supported
       Replicate archive files
       Store archive files
                                                        36
    Archive Files on Jukebox via
    ArchiveLink
   Advantages
       Scalable system
       Long data medium retention period
       Data medium management support
       Jukebox can also be used for other R/3 tasks
   Disadvantages
       Can be a costly solution
       Single document access only feasible for small
        archive files
       Access time heavily depends on archive file size

                                                           37
          ArchiveLink Capabilities
                       With ArchiveLink you can:
                               Archive incoming R/3 documents
                               Archive information created in R/3
                               Directly access the optically archived
R/3 System                      documents from the R/3 applications
                               Integrate an optical archive system
                                with R/3
    ArchiveLink
                               Replace your existing paper archive
                                with electronic archive systems




Archive
system
                                                                 38
    Some Considerations for Media

   Jukebox via ArchiveLink
      Archived data seldom accessed individually

      Access time is not critical, random access slow in Rel. < 3.1G

   HSM system with block access
      Archived data are often accessed individually

      Access time can be on the order of 30 seconds

   Hard disk
      Archived data are regularly accessed individually

      Access time must be on the order of 5 seconds

   Magnetic tape
      Archived data are only accessed in exceptional cases




                                                                        39
       Sample Access Times for
       Archived Data
Sec.
300
                                Hard Disk
270
240
210                             HSM
180
150
                                ArchiveLink
120                             random
 90                             access

 60                             ArchiveLink
 30                             reporting

  0
       25 MB   50 MB   100 MB
                                              40
What We’ll Cover

   Archiving benefits
   Definition of terms
   Archiving features
   Archiving steps
   Storage and management
   Archiving by module
   What next?
                             41
FI Documents

   600,000 FI documents per month
   Retention period up to 18 months
   120 GB used space in DB for FI data
   100,000 documents archived in 3 hours,
    80 MB archive file
   Reduce retention period to 3 months
   Data archiving by IT organization

                                       42
MM Documents

   450,000 MM documents and 2,500,000 line
    items per month
   Line-item table bigger than 10 GB
   100,000 line items being archived in 3 hours
   Direct access implemented by customer
   Reduce retention period
   Random access and reporting on data archive
    files shows good performance

                                             43
    Access Times … Example

   Optimal situation, single job, elapsed
    time
    Archiving Object   create 1000   archive 1000
    FI_DOCUMNT         500s          60s
    SD_VBAK            1000s         160s
    SD_VBRK            1000s         140s + 60s
    RV_LIKP            1000s         340s
    MM_MATBEL          1000s         60s + 60s


                                                    44
What We’ll Cover

   Archiving benefits
   Definition of terms
   Archiving features
   Archiving steps
   Storage and management
   Archiving by module
   What Next?
                             45
    What next?

   Increase in third party solutions
   Better reporting for archiving objects
   More robust document imaging solutions
   Enhanced performance of ADK
   Further integration of archiving and
    database administration


                                         46
    Sources of Information

   Check OSS for your platform
   SAPNET good source of information
   SAPLabs, SAP Search, mySAP.com
   SAP’s Early Watch for data archiving
    https://www014.sap-ag.de/dma
   SAP-R3-L@LISTSERV.SAPFAQ.COM


                                           47
Your Turn!         (to launch Q&A)

    PCS Limited, A Professional
      IT Consulting Services
            Company
    “Creating Solutions Through Knowledge”


            PCS Limited
        5425 Parkford Circle
    Granite Bay, CA 95746 USA
         Tel 916.773.9991
         Fax 916.773.9998
        www.pcs-limited.net
    rparkinson@pcs-limited.net
                                             48

								
To top