Implementing the IBM General Parallel File System (GPFS) in a Cross Platform Environment

Implementing the IBM General Parallel File System (GPFS) in a Cross Platform Environment

by Dino Quintero, Matteo Barzaghi, Randy Brewster, Wan Hee Kim, Steve Normann, Paulo Queiroz, Robert Simon, Andrei Vlad
Released June 2011
Publisher(s): IBM Redbooks
ISBN: None

Read it now on the O’Reilly learning platform with a 10-day free trial.

O’Reilly members get unlimited access to books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.

Start your free trial

Book description

This IBM® Redbooks® publication provides a documented deployment model for IBM GPFS™ in a cross-platform environment with IBM Power Systems™, Linux, and Windows servers. With IBM GPFS, customers can have a planned foundation for file systems management for cross-platform access solutions.

This book examines the functional, integration, simplification, and usability changes with GPFS v3.4. It can help the technical teams provide file system management solutions and technical support with GPFS, based on Power Systems virtualized environments for cross-platform file systems management.

The book provides answers to your complex file systems management requirements, helps you maximize file system availability, and provides expert-level documentation to transfer the how-to skills to the worldwide support teams.

The audience for this book is the technical professional (IT consultants, technical support staff, IT architects, and IT specialists) who is responsible for providing file system management solutions and support for cross-platform environments that are based primarily on Power Systems.

Table of contents

  1. Notices
    1. Trademarks
  2. Preface
    1. The team who wrote this book
    2. Now you can become a published author, too!
    3. Comments welcome
    4. Stay connected to IBM Redbooks
  3. Chapter 1: Introduction
    1. Overview and strategy
    2. Features
    3. Licensing
    4. Operating systems support
    5. Hardware support
    6. Contact information
  4. Chapter 2: Infrastructure planning
    1. Network design
      1. The 1 Gigabit Ethernet and 10 Gigabit Ethernet
      2. InfiniBand (IP over IB/RDMA)
      3. Configuring the NIC
      4. IBM data center networking (DCN) products
    2. Storage design
      1. Host bus adapter (HBA)
      2. Multipath driver
      3. The characteristic of storage hardware
      4. Tape library
    3. InfiniBand Storage DCS9900
    4. Industry solution design
      1. High-Performance Computing
      2. RDBMS (1/2)
      3. RDBMS (2/2)
      4. Web application server
      5. Media contents delivery
      6. Commercial industry
    5. GPFS planning guidance
      1. Operating system and file systems with multiplatform GPFS
      2. Security
      3. High availability (1/2)
      4. High availability (2/2)
      5. Network Shared Disk (NSD) creation considerations (1/2)
      6. Network Shared Disk (NSD) creation considerations (2/2)
      7. GPFS considerations (1/2)
      8. GPFS considerations (2/2)
      9. Tivoli Storage Manager for GPFS
    6. Summary
  5. Chapter 3: Scenarios
    1. ITSO laboratory environment
      1. Diagram of ITSO laboratory
      2. ITSO laboratory environment host file
    2. Three-node GPFS cluster using internal disks
      1. Requirements: Hardware, software, network, storage
      2. GPFS configuration
      3. Diagram of GPFS three-node cluster with internal disks
      4. Setting up and configuring a three-node cluster
    3. Recovery on the three-node cluster
    4. Linux InfiniBand cluster with RDMA and Linux for System x clients
      1. Requirements: Hardware and software
      2. Ethernet GPFS cluster
      3. InfiniBand GPFS cluster
    5. Cross-platform cluster: Windows servers, Linux/AIX clients
      1. Requirements: Hardware, software, network, storage
      2. GPFS cluster configuration
      3. GPFS cluster diagram
      4. Installing Windows Server 2008 R2
      5. Install the IBM GPFS Windows version
      6. GPFS Windows cluster setup (1/2)
      7. GPFS Windows cluster setup (2/2)
      8. Simulating a GPFS failure on the cluster
      9. Adding x86 Linux node to the Windows cluster
    6. DB2 pureScale InfiniBand cluster on AIX
      1. Requirements: Hardware, software, network, storage
      2. GPFS configuration
      3. GPFS cluster diagram
      4. Cluster setup and configuration
      5. Use the mmpmon tool to collect statistics from each node
      6. Simulating a GPFS failure on member 0 (algeria)
      7. Troubleshooting DB2 pureScale issues
    7. Multi-cluster configuration
      1. Requirements: Hardware, software, network, storage
      2. GPFS configuration
      3. GPFS multi-cluster diagram
      4. Multi-cluster setup and configuration
      5. Multi-cluster scenario
      6. Add security at the GPFS communication network on Bronx cluster
      7. Verify GPFS daemon is active on all nodes of both clusters
    8. Disaster recovery using GPFS replication
      1. Requirements: Hardware, software, network, storage
      2. GPFS configuration
      3. GPFS configuration diagram
      4. Set up and configure GPFS DR cluster (1/5)
      5. Set up and configure GPFS DR cluster (2/5)
      6. Set up and configure GPFS DR cluster (3/5)
      7. Set up and configure GPFS DR cluster (4/5)
      8. Set up and configure GPFS DR cluster (5/5)
  6. Chapter 4: Management and maintenance
    1. Migration and update
      1. Migrating GPFS from 3.2 or 3.3 to 3.4
      2. Applying corrective fixes to GPFS
      3. Migrating data to new storage
      4. Reinstalling a node (mmsrdfs)
    2. Managing a GPFS cluster
      1. Adding a node to the cluster
      2. Adding a disk to a file system
      3. GPFS network usage
      4. Adding a remote cluster
      5. Adding or removing a file system
      6. Enabling GPFS replication
      7. Exporting or importing a file system
    3. Reducing file system fragmentation: The mmdefragfs command
    4. Optimizing extended attributes: The fastea option
    5. Setting up GPFS callback functionality: The callback commands
    6. Monitoring GPFS configuration status: The SNMPD protocol
    7. SSH configuration
  7. Chapter 5: Performance and tuning
    1. GPFS architecture
    2. NSD considerations
    3. GPFS parametrization
      1. File system parametrization
      2. Block allocation methods
      3. GPFS general parametrization
    4. Network parametrization (1/2)
    5. Network parametrization (2/2)
    6. Parametrization
      1. Controlling the bandwidth
      2. Controlling GPFS parallelism
      3. Linux considerations
      4. Storage device considerations
    7. Monitoring performance
  8. Chapter 6: Problem determination
    1. Problem determination process
      1. Defining the problem
      2. Gathering information from the user
      3. Gathering application, GPFS, and system information
    2. GPFS debug commands
      1. Data collection commands
      2. Data analysis commands and scripts
      3. mmfs logs
    3. GPFS problem scenarios
      1. Considerations
      2. Scenario 1: Upgrading GPFS
      3. Scenario 2: Analyzing waiters
      4. Scenario 3: Application failure
      5. Scenario 4: GPFS file system hang
      6. Scenario 5: File system unmounting
  9. Chapter 7: IBM Power Systems virtualization and GPFS
    1. IBM Power Systems (System p)
      1. Introduction
      2. Virtual I/O Server (VIOS)
      3. VSCSI and NPIV
      4. Virtual SCSI target adapters (VSCSI)
      5. N_Port ID Virtualization (NPIV) (1/2)
      6. N_Port ID Virtualization (NPIV) (2/2)
    2. Shared Ethernet Adapter and Host Ethernet Adapter
      1. Shared Ethernet Adapter (SEA)
      2. Host Ethernet Adapter (HEA)
  10. Chapter 8: Information lifecycle management (ILM)
    1. Explaining the ILM concept
      1. Snapshot management tasks
      2. Storage pools
      3. File sets
      4. Policies and rules
      5. ILM data flow
    2. Tivoli Storage Manager
      1. Preparing the Tivoli Storage Manager server
      2. Tivoli Storage Manager installation (1/2)
      3. Tivoli Storage Manager installation (2/2)
      4. Administrator Center installation
      5. Tivoli Storage Manager server configuration
      6. Tivoli Storage Manager client configuration
    3. Information lifecycle management scenarios
      1. Working with snapshots
      2. Working with storage pools and policies (1/2)
      3. Working with storage pools and policies (2/2)
      4. Working with external storage pool and GPFS policies
      5. Working with hierarchical storage management
      6. Working with hierarchical storage management and GPFS policies
      7. End-to-end implementation of ILM and HSM
    4. Backup and restore
      1. GPFS backup tools
      2. GPFS advanced backup tools
      3. Conclusions
    5. Advanced external pool scenario
      1. Off-line compressed storage
      2. Implementation scenario
  11. Chapter 9: Disaster recovery using GPFS
    1. Disaster recovery solution using GPFS replication
      1. Configuration
      2. Characteristics of this DR configuration
    2. The GPFS mmfsctl command
  12. Related publications
    1. IBM Redbooks
    2. Other publications
    3. Online resources
    4. How to get Redbooks
    5. Help from IBM
  13. Index (1/3)
  14. Index (2/3)
  15. Index (3/3)
  16. Back cover

Product information

  • Title: Implementing the IBM General Parallel File System (GPFS) in a Cross Platform Environment
  • Author(s): Dino Quintero, Matteo Barzaghi, Randy Brewster, Wan Hee Kim, Steve Normann, Paulo Queiroz, Robert Simon, Andrei Vlad
  • Release date: June 2011
  • Publisher(s): IBM Redbooks
  • ISBN: None