Measuring backend capacity in primary server storage pools.

Technote (FAQ)


Question

I have Tivoli Storage Manager Suite for Unified Recovery. What tools are available to help me measure capacity in Tivoli Storage Manager primary storage pools and Tivoli Storage Manager FastBack repositories?

Answer


Introduction

Tivoli Storage Manager Suite for Unified Recovery is a bundle of Tivoli Storage Manager products that offers simplified pricing and licensing with a tiered per-terabyte (TB) metric. You can install as many of the individual products as needed for your environment. There is no per-server, per-installation or processor value unit (PVU) charge; you are only charged for the TB that is stored in Tivoli Storage Manager primary storage pools and FastBack repositories.

This document is organized into a number of sections. The sections are:

  • Section 1 provides planning information to consider prior to performing capacity measurements on one or more TSM servers.
  • Sections 2 – 4 are instructions on how to collect the capacity measurement data for various TSM server versions.
  • Section 5 provides information the capacity measurement program and details about what is measured and what is excluded.

Consider the following table to determine which sections to refer to in order to collect capacity measurement information for your environment:

Server Version
Considerations
Use Section
6.2
2
6.3
3
7.1
3
7.1.1
  • Server to server replication is not being used.
  • Server to server replication is being used and the source and target servers are using the same policies.
3
7.1.1
  • Server to server replication is being used and the source and target servers are using different policy settings.
4

Note: Versions V5.5 or V6.1 are no longer supported. For information on how capacity measurements were performed for these versions of TSM, refer to:

http://www-01.ibm.com/support/docview.wss?uid=swg21684142.


Example Scenario:
For your environment you have 8 TSM servers. The details about those servers are:

  • 5 servers are V6.3.5
  • 1 server is 7.1.0
  • 2 servers are 7.1.1, they use TSM server to server replication, and the target server is managing the replicated data using different policies than the source server.

Referring to the table above, the instructions in section 3 would be used for the five 6.3.5 servers and the one 7.1.0 server. The instructions in section 4 would be used for the two 7.1.1 servers that are using server to server replication and using different policies to manage the replicated data.

Overview of collecting capacity measurements

Sections 2, 3, and 4 provide the specific details of how to perform a capacity measurement for the various server versions. The general approach for each section is:
1. Download the appropriate macro for the server version where the measurement is being performed.

2. Invoke the macro from a command line administrative client.

  • It is possible to do this from a command line administrative client running on the actual server machine. This requires that you logon to the server machine along with having an appropriate TSM administrative ID to use when you invoke the command line administrative client.
  • It is also possible to do this from a single machine where a TSM command line administrative client is installed. From here, you would just connect through TCP/IP to each TSM server.

3. Each server measured will generate an output report. Within each, it will report the measured Terabyte (TB) capacity value to consider. Each report will report the capacity value with the following stanza:

Data Managed by TSM Suite for Unified Recovery
TB within PRIMARY storage pools:  NNN

Where “NNN” is the Terabyte (TB) capacity value to use for the measurement.

4. The values measured for all applicable TSM servers should than be summed in order to determine the SUR capacity measurement value.

Section 1: Planning for capacity measurements

Consider the following prior to collecting capacity measurements on your TSM servers.

  • Administrative SQL commands require CPU, memory, and I/O to the database for the TSM server. Generally, for V6 and V7 servers, these requirements are not an issue as the improved SQL processing engine provided by DB2 handles these without an issue. For V6 and V7 servers, running the macro usually takes many minutes to tens of minutes depending upon the size of the server and the data stored on that server.
  • Consider the current health of the TSM server. If the server is I/O constrained to the database or it is having difficulty supporting existing client or daily maintenance workloads, than running the capacity measurement macro may take a long time to complete or it may appear to hang or never complete.
  • The SUR macro applies exclusions to the data that it is measuring. As such, the TB capacity value reported by the macro will not match or align directly with the results from QUERY OCCUPANCY or QUERY AUDITOCCUPANCY or their equivalent SQL statements. This is not a problem.
  • TSM Version 6 server users that have enabled server deduplication or client deduplication may be affected by code defects that cause the occupancy values to be incorrect. The following two APAR's (below) should be reviewed to determine if these apply to your environment. If one or the other does apply to your environment, apply the corresponding server fix-pack or patch containing the fix for this APAR along with performing any additional corrective actions described in the APAR's or any corresponding techdocs that reference these APARs. After these fixes have been applied and corrective actions performed, the capacity values used for the measurement will no longer be affected by these.

    IC73005 - Available, or targeted to be available, in fix pack level 6.2.3, http://www.ibm.com/support/docview.wss?uid=swg1IC73005.
    IC73592 – Available, or targeted to be available, in fix packs 6.1.5, 6.2.3 and patch level 6.2.2.3, http://www.ibm.com/support/docview.wss?uid=swg1IC73592.


Section 2: Measuring capacity for a 6.2 server

To measure capacity for a 6.2 server:

1. Download the macro tsm_tb_cap_v6.macro. This will download automatically from:

TSM_V6_Macro

2. For each TSM server, run the following command:

dsmadmc –id
=<adminid> -password =<password> -dataonly=yes –outfile =<output_file_name> macro tsm_tb_cap_v6.macro

The input parameters to provide for this command are:
  • <adminid> is the ID for a registered administrator to the TSM server that you are collecting information from.
  • <password> is the password for the administrator ID being used.
  • <output_file_name> is the name of the file that will contain the output from this measurement collection. Note that if you are performing measurements for many TSM servers, the name used for this file should include the server name in it. For example, if this measurement is being run against server DC1SRV100, than consider an output file name of “dc1srv100.surcap.out”.


Section 3: Measuring capacity for a 6.3, 7.1, or 7.1.1 server

This section covers how to measure capacity for server versions:

  • 6.3 or 7.1
  • 7.1.1 servers that are not using TSM server to server replication.
  • 7.1.1 servers that are using TSM server to server replication and using the same version and retention policies.

To measure capacity for a 6.3, 7.1, or 7.1.1 server:
1. Download the macro tsm_tb_cap_v63.macro. This will download automatically from:

TSM_V63_Macro

2. For each TSM server, run the following command:

dsmadmc –id
=<adminid> -password =<password> -dataonly=yes –outfile =<output_file_name> macro tsm_tb_cap_v63.macro

The input parameters to provide for this command are:
  • <adminid> is the ID for a registered administrator to the TSM server that you are collecting information from.
  • <password> is the password for the administrator ID being used.
  • <output_file_name> is the name of the file that will contain the output from this measurement collection. Note that if you are performing measurements for many TSM servers, the name used for this file should include the server name in it. For example, if this measurement is being run against server DC1SRV100, than consider an output file name of “dc1srv100.surcap.out”.

Note that for 7.1.1 servers, the 7.1.1 version of the Operations Center will collect this capacity measurement automatically. The capacity measurement information is available in the “SUR Occupancy” column on the “Servers” page in the interface. This also shows when the data was collected in the “SUR Collected” column on this page. To use this, the Operations Center needs to be installed and configured for your environment. It also requires that the hub server be version 7.1.1 and any corresponding spoke servers where you want the SUR measurement data automatically collected to also be at version 7.1.1.


Section 4: Measuring capacity for a 7.1.1 server that manages TSM replicated data using different policies

This section covers how to measure capacity for 7.1.1 servers that:

  • Perform server to server replication
  • The replicated data between the source and target server is managed using different policies.
  • To determine if different policies are used to manage replicated data, issue the command “QUERY REPLSERVER” on either the source or the target server. A server is managing replicated data using different policies if this query result reports “Dissimilar Policies: On”.

To measure capacity for a 7.1.1 server that manages replicated data using different policies:


1. Download the macro tsm_tb_cap_v711_src.macro. This will download automatically from:

TSM_V711_Src.macro

2. Download the macro tsm_tb_cap_v711_tgt.macro. This will download automatically from:

TSM_V711_Tgt.macro

3. For each source TSM server, run the following command:

dsmadmc –id
=<adminid> -password=<password> -dataonly=yes –outfile=<output_file_name> macro tsm_tb_cap_v711_src.macro

The input parameters to provide for this command are:

  • <adminid> is the ID for a registered administrator to the TSM server that you are collecting information from.
  • <password> is the password for the administrator ID being used.
  • <output_file_name> is the name of the file that will contain the output from this measurement collection. Note that if you are performing measurements for many TSM servers, the name used for this file should include the server name in it. For example, if this measurement is being run against server DC1SRV100, than consider an output file name of “dc1srv100src.surcap.out”.


4. For each target TSM server, run the following command:

dsmadmc –id
=<adminid> -password =<password> -dataonly=yes –outfile =<output_file_name> macro tsm_tb_cap_v711_tgt.macro

The input parameters to provide for this command are:
  • <adminid> is the ID for a registered administrator to the TSM server that you are collecting information from.
  • <password> is the password for the administrator ID being used.
  • <output_file_name> is the name of the file that will contain the output from this measurement collection. Note that if you are performing measurements for many TSM servers, the name used for this file should include the server name in it. For example, if this measurement is being run against server DC1SRV100, than consider an output file name of “dc1srv100tgt.surcap.out”.
5. For the SUR capacity measurement, compare the results from the source and target servers and use the larger of the two values.
  • An example result from a source server is:

    ***********************************************************
    **** Tivoli Storage Manager Suite for Unified Recovery ****
    ****           Terabyte (TB) Capacity Report           ****
    ****           for Use with V7.1.1 or higher           ****
    ****                                                   ****
    ****       Special handling for source server using    ****
    ****        dissimilar policies with replication.      ****
    ****************** (macro version: 2.0) *******************

    Capacity Report for server:  DC1SRV100

    Report generatation date:  Aug 25, 2014

    Data Managed by TSM Suite for Unified Recovery
    TB within PRIMARY storage pools:  155

    *----------------- Deduplication Benefits ----------------*
    TSM Data deduplication resulted in
     TB being excluded from measurement:  32

  • An example result from a target server is:

    ***********************************************************
    **** Tivoli Storage Manager Suite for Unified Recovery ****
    ****           Terabyte (TB) Capacity Report           ****
    ****           for Use with V7.1.1 or higher           ****
    ****                                                   ****
    ****       Special handling for target server using    ****
    ****        dissimilar policies with replication.      ****
    ****************** (macro version: 2.0) *******************

    Capacity Report for server:  DC2SRV010T

    Report generatation date:  Aug 25, 2014

    Data Managed by TSM Suite for Unified Recovery
    TB within PRIMARY storage pools:  132

    *----------------- Deduplication Benefits ----------------*
    TSM Data deduplication resulted in
     TB being excluded from measurement:  45

For this example, the two capacity values reported are 155 for the source and 132 for the target. In this case, the value to use for the capacity measurement is 155.


Section 5: Background

Tivoli Storage Manager Suite for Unified Recovery delivers a comprehensive set of data protection capabilities such as:

  • Massive scalability with broad platform support
  • Target-side and source-side data deduplication
  • Advanced support for virtual environment protection
  • Online, consistent, and centralized data protection for:
    * Databases
    * mySAP/SAP R3 environments
    * Email servers running IBM Lotus Domino or Microsoft Exchange
    * Item-level recovery of Microsoft Exchange email objects
  • LAN-free backup and restore operations that utilizes the SAN to remove data transfer from the LAN, providing high-performance backup and restore processing while helping to minimize network traffic
  • File-system and application-aware snapshots and Continuous Data Protection (CDP)
  • Bare machine recovery for Microsoft Windows and Linux servers, for servers that are already protected with Tivoli Storage Manager FastBack
  • Hierarchical storage management for Unix and Linux systems

The following products comprise Tivoli Storage Manager Suite for Unified Recovery solution:

IBM Tivoli Storage Manager Extended Edition: Use this product for highly-scalable, enterprise-class backup, restore, and archive processing, and disaster recovery

IBM Tivoli Storage Manager for Databases: Use this product for nondisruptive protection of Oracle and Microsoft SQL data

IBM Tivoli Storage Manager for Enterprise Resource Planning: Use this product to protect vital SAP R/3 system data more efficiently, consistently and reliably

IBM Tivoli Storage Manager for Mail: Use this product to secure Lotus Domino and Exchange data, and to provide granular restore processing Exchange objects

IBM Tivoli Storage Manager for Storage Area Networks: Use this product to maximize storage network connections for Tivoli Storage Manager servers and client computers

IBM Tivoli Storage Manager for Space Management: Use this product to move inactive data to reclaim online disk space

IBM Tivoli Storage Manager for Virtual Environments: Use this product to protect and provide flexible recovery of VMware environments

IBM Tivoli Storage Manager FastBack: Use this product to provide advanced data protection and near-instant recovery for Windows and Linux servers

IBM Tivoli Storage Manager FastBack for Bare Machine Recovery: Use this product to restore a Windows or Linux Server Operating System volume within an hour

IBM Tivoli Storage Manager FastBack for Microsoft Exchange: Use this product for fast and easy recovery of individual e-mail objects

For information on supported levels of bundled products see this link:

http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?subtype=ca&infotype=an&supplier=897&letternum=ENUS211-201

Frequently Asked Questions

Please refer to the FAQ technote for hints, tips, guidance for performing a capacity measurement. The FAQ technote is: http://www-01.ibm.com/support/docview.wss?uid=swg21656476.

Measuring capacity

Administrative macro commands are available to help you measure TB stored in Tivoli Storage Manager Primary Storage Pools. There are also instructions available that explain how to us the Tivoli Storage Manager FastBack graphical user interface to help you measure TB stored in Tivoli Storage Manager FastBack Repositories

How data is measured

The Administrative macro commands use the following rules when measuring data:

  • Measures the amount of deduplicated data in a deduplicated primary storage pool, where Tivoli Storage Manager has deduplicated data. This measurement shows the cost efficiency of using Tivoli Storage Manager data deduplication when the number of TB in primary storage pools is reduced.

  • Excludes copy pool and active data pools. These are copies of primary storage pools and not measured for capacity reporting purposes.

  • Excludes data from Tivoli Storage Manager HSM for Windows and Tivoli Storage Manager for SharePoint. These OEM products are sold on a Processor Value Unit (PVU) basis and are not included in the Tivoli Storage Manager Suite for Unified Recovery.

  • Measures virtual volume source data (excluding target data). TSM allows a server (a source server) to store the results of database backups, export operations, storage pool operations, and a DRM PREPARE command on another TSM server (a target server). The data is stored as virtual volumes. When querying TSM, virtual volume data appears in primary storage pools in both the target and the source server. The data that is stored on the target server is a Disaster Recovery (DR) copy of data and is not measured.

  • Excludes data replicated by FastBack to Tivoli Storage Manager and excludes data that was backed up from a FastBack server using a Tivoli Storage Manager client. This data represents a copy of data on the FastBack server that is already measured.

  • Excludes other duplicated data. Other duplicated data represents data that is reported in primary storage pools, which is a copy of data that is already stored by other means. For example, the table of contents for a backup set is stored in a primary storage pool. The table of contents is metadata that describes the data on a backup set so this data is excluded from the measurement.

  • Excludes data from client nodes configured as REPL_MODE of RECEIVE or SYNCRECEIVE on the target server where TSM is configured to perform server to server replication. Note that this exclusion only applies if TSM server to server replication is used and the data is managed by the same policies. See section 4 for an explanation of how replicated data is measured if different policies are used to manage the data.

Measuring data in FastBack repositories


You can measure data within Tivoli Storage Manager FastBack repositories by using the graphical user interface for TSM FastBack. To learn how to use the FastBack user interface to measure capacity in FastBack repositories, see the following web page:


ftp://public.dhe.ibm.com/storage/tivoli-storage-management/capacity_measurement_tools/Measuring_data_stored_in_TSM_FastBack_repositories.pdf

Rate this page:

(0 users)Average rating

Add comments

Document information


More support for:

Tivoli Storage Manager
Server

Software version:

6.2, 6.3, 7.1, 7.1.1, All Supported Versions

Operating system(s):

AIX, HP-UX, Linux, Solaris, Windows

Software edition:

All Editions

Reference #:

1500482

Modified date:

2011-05-27

Translate my page

Machine Translation

Content navigation