Skip to main content

Autonomic features of the IBM Virtualization Engine

Lori Simcox, Advisory Software Engineer, IBM
Lori Simcox is an Advisory Software Engineer in Tivoli Storage software development at IBM in San Jose, California. She received a B.S. degree in Mathematics from Pennsylvania State University and a M.S. degree in Computer Science from the University of Southern California. For the past eight years she has worked on IBM Tivoli Storage Manager and IBM TotalStorage Productivity Center. Her interests also include graphical user interface development and usability.

Summary:  Learn about the many autonomic capabilities of the IBM® Virtualization Engine™ and see how they improve the availability of resources and the efficiency of systems and storage administrators. This sophisticated suite of products manages the servers, storage, systems, and networks across a cross-platform distributed IT environment. It is a key component of an on demand solution to optimize the management of your infrastructure according to your business goals.

Date:  14 Sep 2004
Level:  Introductory
Comments:  

Introduction

Autonomic computing technology encompasses any feature that enables software to become self-managing, self-healing, self-configuring, self-optimizing, and self-protecting in order to reduce management tasks and human error. Such features include, though are not limited to, automated problem determination, common system administration, automatic resource management, provisioning, and orchestration. Just as the body's autonomic nervous system regulates such functions as heart rate, blood pressure, and respiration for a person's health, autonomic computing systems control the operation, optimization, and availability of the systems and storage in a distributed IT infrastructure.


What is the IBM Virtualization Engine

The IBM Virtualization Engine is a suite of systems services and technologies that provides a comprehensive approach to managing resources, servers, storage, and networks as an integrated system, rather than as individual components. It reduces costs, improves storage utilization, and increases the amount of managed resources per administrator. The IBM Virtualization Engine is a key part of an on demand operating solution in that it helps you optimally manage your IT environment in a simpler, more efficient, and more responsive manner to meet the goals of your business. The IBM Virtualization Engine provides you with a high-level view of the computing resources that make up your enterprise; by doing so, it hides the details of the underlying diverse systems and hardware. It includes powerful tools for automating systems provisioning, workload management, storage virtualization, and partitioning. The IBM Virtualization Engine consists of a set of systems technologies and a set of systems services that provide a consistent logical view of your cross-platform IT environment.

The IBM Virtualization Engine Systems Technologies for eServer™ servers includes Hypervisor™, which helps partition and move resources dynamically in multiplatform environments, virtual Ethernet, which allows the efficient prioritization of traffic on shared networks, and virtual I/O, which allows flexible allocation, management, and sharing of physical resources (storage, adapters, and devices) among multiple partitions. The Systems Technologies also provide simultaneous multithreading, dynamic logical partitioning, and uncapped partitions for efficiently sharing server processor resources.

The IBM Virtualization Engine Systems Services consists of services for eServers and for storage. The IBM Virtualization Engine System Services for eServers uses IBM's Integrated Solutions Console as its user-friendly, dashboard-like view of the status of your environment. The IBM Integrated Solutions Console is a portal-based console that lets you keep track of the systems and resources in your environment by allowing common systems administration across product boundaries.

The IBM Virtualization Engine Suite for Servers consists of the following products and tools:

  • IBM Tivoli® Provisioning Manager
  • IBM Grid Toolbox
  • IBM Enterprise Workload Manager
  • IBM Director Multiplatform

The IBM Virtualization Engine for Storage consists of the following, delivered as separate products:

  • IBM TotalStorage® Productivity Center
  • IBM TotalStorage SAN Volume Controller
  • IBM TotalStorage SAN File System

This article discusses the autonomic computing capabilities of the IBM Virtualization Engine's Systems Services and describes their value to IT administrators in managing their enterprise systems and storage.


IBM Tivoli Provisioning Manager for IBM eServer

IBM Tivoli Provisioning Manager brings intelligent provisioning of servers, operating systems, middleware, applications, storage, and network devices to the IBM Virtualization Engine suite. Provisioning occurs as a series of ordered steps. The steps might include anything from installing operating systems, configuring networks (VPNs), setting up storage environments (such as disk, SAN, or NAS), authenticating users or applications, and allocating resources to and from the available pool. IBM Tivoli Provisioning Manager automates the manual tasks of provisioning, configuration, and deployment by using built-in workflows. The workflows perform the steps needed to deploy the vendors' hardware and systems. IBM Tivoli Provisioning Manager also contains a workflow editor that allows you to create workflows tailored to the automation needs of your business. By automating numerous tasks an administrator must do, IBM Tivoli Provisioning Manager decreases the potential for human error and increases the amount of resources an administrator can manage.


Figure 1. IBM Tivoli Provisioning Manager showing storage workflows
IBM Tivoli Provisioning Manager

IBM Grid Toolbox

The IBM Grid Toolbox provides a powerful set of building block tools to create, deploy, and administer services in a grid. A grid is a potentially vast group of distributed, cross-platform computational resources connected by an infrastructure adhering to the Open Grid Services Infrastructure (OGSI) standards. IBM Grid Toolbox allows the systems administrator to more efficiently and more rapidly develop applications and processes for a grid. The multiplatform hosting environment provides tools to effectively manage common grid services such as container management, logging, security, resource management, and data management. It also provides Common Management Model, Policy and IBM Service Group services, which are unique to the IBM Grid Toolbox.


IBM Enterprise Workload Manager

IBM Enterprise Workload Manager (EWLM) allows administrators to coordinate the activities of a large number of cross-platform servers to optimize performance and to meet overall business goals. EWLM takes the mature technology and goal-oriented philosophy of workload management on z/OS® and applies it to the distributed environment. Examples of workload management decisions are diverting work from systems that are down, averting work from systems with a heavy load to those with a lighter load, and sending work to systems that are running work of lesser importance as determined by business policies. EWLM's Web-enabled Control Center provides a user-friendly Policy Wizard to define service policies (normal and weekend, for example), to specify service class goals (average response time, percentile response time, velocity, and so on), and to define policies for transactions and processes. EWLM's domain manager deploys the business policy to the managed platforms and receives performance data back at regular intervals, which enable it to show an end-to-end view.

Application specific data is collected using the industry standard Application Response Measurement (ARM) implementations. The EWLM Control Center allows administrators to understand the quality of service delivery by displaying topologies with drill-downs to application and server levels, response times, server resource utilization, and server delays. Such information, along with process and transaction details, gives administrators a good understanding of service delivery from a customer's perspective by identifying bottlenecks and associated server resource consumption. IBM Enterprise Workload Manager provides the self-optimization attribute of autonomic computing technology by providing information about performance problems, the components contributing to those problems, and the workloads in the infrastructure impacted by the problem. EWLM's self-tuning features are necessary to ensure the system and application availability customers expect, as well as the optimal performance of the system load that keeps a business running according to its goals.


Figure 2. IBM Enterprise Workload Manager showing transaction classes
IBM Enterprise Workload Manager

IBM Director Multiplatform

Often, administrators manage AIX®, Linux, and Windows® servers as individual physical servers. By consolidating to IBM eServer BladeCenter™ technology, companies can realize immediate benefits; however, the problem of managing the respective AIX, Linux, and Windows servers and populating new blades still exists. IBM Director Multiplatform unites those disparate operating systems into a single, user-friendly view. It automates such systems management tasks as resource monitoring, task health, corrective management, and console launching. Its inventory functions discover all the hardware available in the environment, its configuration, and its status by scanning each managed system. Automated problem determination capabilities point out problems with hardware, such as faulty power supplies, fans, voltage regulator modules, and network interface cards, among others. Administrators can monitor specific processes and system resources and receive alerts when stated thresholds are reached or when a process does not start. IBM Director Multiplatform allows administrators to define event triggers separately from actions, providing the capability to create custom action plans for various systems and groups of managed systems. Its mass configuration capability can configure SNMP community names and trap destinations for multiple systems at once. IBM Director Multiplatform can easily deploy asset identification across machines and remotely perform network operations such as specifying domains, adding DNS servers to network properties, and setting DHCP on managed systems. Through its remote management, automated problem determination, and self-monitoring autonomic computing capabilities, IBM Director Multiplatform increases the availability of managed servers as well as the efficiency of administrators by reducing the down time of hardware and managed systems.


IBM TotalStorage Productivity Center

IBM TotalStorage Productivity Center manages the storage, devices, and SAN fabric to aid in storage planning, capacity utilization, performance, and storage and application availability of a distributed, cross-platform environment. It provides a user interface into IBM Tivoli Storage Resource Manager, IBM Tivoli SAN Manager, and IBM TotalStorage Multiple Device Manager. The IBM TotalStorage Productivity Center with Advanced Provisioning provides a set of Tivoli Provisioning Manager storage workflows that automate many of the operations associated with provisioning of SAN-attached storage. IBM TotalStorage Productivity Center provides self-optimizing, self-healing, self-managing, and automated problem determination autonomic computing capabilities.


Figure 3. IBM TotalStorage Productivity Center
IBM TotalStorage Productivity Center

IBM Tivoli Storage Resource Manager

IBM Tivoli Storage Resource Manager (SRM) helps storage administrators wisely manage their storage budgets by using existing storage more efficiently, allowing them to better predict future storage needs, and enabling them to keep data available to applications. Discovery and scanning capabilities by the lightweight IBM Tivoli SRM agents automatically monitor disks, storage subsystems, partitions, shared directories, and host or SAN-attached storage. They help administrators keep track of capacity and consumption used by file systems, directories, files, and users. IBM Tivoli SRM provides over 300 customizable reports, which give details about file systems, databases, assets, availability, capacity, backup, usage, and usage violations within the storage infrastructure. Administrators can define alerts on physical objects (disks, partitions, servers, for example) or logical entities (such as file systems, files, databases, tables). Using these alerts, which specify an action to be performed if a given event occurs or if a specific condition is met, IBM Tivoli SRM provides self-monitoring and self-healing capabilities across the storage environment. As add-ons to IBM Tivoli Storage Resource Manager, IBM Tivoli SRM for Databases reports on storage utilization for application databases, finds unused space within the database tables, and identifies the fastest growing databases in the infrastructure. IBM Tivoli SRM for Chargeback generates reports for charge back for storage usage. IBM Tivoli Storage Resource Manager's enterprise-wide reporting and its self-monitoring capabilities help administrators intelligently plan storage capacity, improve storage utilization, and ensure the availability of applications and storage.


Figure 4. IBM Tivoli Storage Resource Manager
IBM Tivoli Storage Resource Manager

IBM Tivoli SAN Manager

Storage administrators can identify faults and resolve problems quickly with the self-monitoring capabilities of IBM Tivoli SAN Manager. IBM Tivoli SAN Manager automatically discovers hosts and devices in your storage network. It depicts your infrastructure as a topology, enabling administrators to drill-down into SANs and devices. Problems are isolated and reported at the exact SAN link at which they occur and up through higher levels of the topology so that problems are immediately apparent. IBM Tivoli SAN Manager discovers multivendor Fibre Channel or iSCSI devices through the use of inband agents (on managed hosts) and outband agents (through SNMP). It queries switches and hosts for other information such as host-level storage, LUNs and file systems, events, and other details available from the HBAs. From a single console, administrators can manage their entire SAN fabric by adding and removing devices from zones. Through IBM Tivoli SAN Manager's proactive monitoring and automated problem determination, storage administrators can ensure uninterrupted data access across their SANs.


Figure 5. IBM Tivoli SAN Manager
IBM Tivoli SAN Manager

IBM TotalStorage Multiple Device Manager

IBM TotalStorage Multiple Device Manager helps storage administrators optimize the configuration, replication, and performance of their SAN storage. IBM TotalStorage Multiple Device Manager automatically discovers Storage Management Interface Specification- (SMI-S) compliant storage devices connected to the SAN. It provides a central view of the many devices on the SAN through the IBM Director Console, which allows the storage administrator to efficiently perform maintenance tasks. Performance thresholds, such as I/O rate, disk utilization, average cache hold time, virtual disk I/O rate, managed disk I/O rate, and percentage of sequential I/Os can be set for each storage device. When a threshold is reached, IBM TotalStorage Multiple Device Manager issues triggered events (set up by the administrator) to self-correct a problem before it becomes critical. Recommended values for thresholds (if they are defined) for IBM storage devices are provided. IBM TotalStorage Multiple Device Manager's Performance Manager performs automated allocation optimization based upon analysis of historical performance statistics to determine the storage resource on the SAN that has the best performance characteristics. When an administrator adds a new LUN to an enterprise storage subsystem, the best LUN is automatically selected because of this self-optimization. The Replication Manager simplifies replication by providing efficient management of Peer-to-Peer Remote Copy and IBM FlashCopy services. Single console management, self-healing, and self-monitoring of SAN resources are important autonomic computing characteristics of IBM TotalStorage Multiple Device Manager.


Figure 6. IBM Director Console displaying IBM TotalStorage Multiple Device Manager
IBM Director Console

IBM TotalStorage® SAN Volume Controller

IBM TotalStorage SAN Volume Controller provides advanced copy, migration, and management capabilities for your distributed storage environment. From a single console, IBM TotalStorage SAN Volume Controller allows administrators to create pools of virtual volumes that can span physical storage resources (volume controllers and their physical volumes). Administrators can create virtual disk groups based on target performance, availability, and cost. As the storage needs of your business increase, IBM TotalStorage SAN Volume Controller provides high performance and continuous availability of storage resources. Its advanced migration functions allow storage capacity to be reallocated and data to be moved from one storage device to another without taking storage or applications offline. Its efficient copy capabilities allow FlashCopy and Peer-To-Peer Remote Copy to be run across the storage subsystems of different vendors. IBM TotalStorage SAN Volume Controller increases administrator productivity by helping optimize how storage resources are used, viewed, and centrally managed to eliminate overloading and underutilization.


Figure 7. IBM TotalStorage SAN Volume Controller listing virtual disks
IBM TotalStorage SAN Volume Controller

IBM TotalStorage SAN File System

IBM TotalStorage SAN File System allows administrators to dynamically manage and apportion resources based on user-defined policies. It virtualizes storage resources by providing a global namespace within a distributed infrastructure, allowing easier management, access, and capacity planning. A storage administrator sets up policies for the storage pools and the rules for determining where to locate newly created files. IBM TotalStorage SAN File System uses those policies to automatically put new files on the correct storage resource and alerts the administrator when the storage pools reach their quotas. FlashCopy functionality allows backup and recovery of all or part of the file system to reduce disruption during those operations. By the automation of the routine tasks of storage management, IBM TotalStorage SAN File System reduces administrative overhead and provides dynamic allocation and monitoring to your storage environment.


Conclusion

The autonomic computing features of the IBM Virtualization Engine reduce management tasks for systems administrators, increase availability of storage and applications, and lessen human error. In the IBM Virtualization Engine Suite for Servers, IBM Director Multiplatform centrally manages and self-monitors the hardware resources in your diverse IT environment in a homogeneous way. IBM Director Multiplatform works in conjunction with the Virtualization Engine Console to provide the common user interface into your distributed systems and with the IBM Tivoli Provisioning Manager to automatically provision your environment with the necessary applications and storage. IBM Enterprise Workload Manager autonomically optimizes and tunes your network according to business goals to improve the performance of your infrastructure. In the IBM Virtualization Engine Suite for Storage, IBM TotalStorage Productivity Center with its IBM Tivoli Storage Resource Manager, IBM Tivoli SAN Manager, and IBM TotalStorage Multiple Device Manager, autonomically monitors the storage and devices in your cross-platform distributed environment. It automatically detects problems and provides self-healing capabilities through triggered alerts. IBM TotalStorage SAN Volume Controller centralizes SAN management and IBM TotalStorage SAN File System supports self-monitoring and policy-based automation in placing files.

The IBM Virtualization Engine reduces costs, improves storage utilization, strengthens performance, and increases the managed storage and systems per administrator. Just as the body's heart, lungs, and nervous system function autonomically, so does IBM's Virtualization Engine autonomically provision, manage, and administer your IT environment for the smooth operation of the network, servers, applications, and storage. IBM's self-optimizing, self-managing, and self-healing Virtualization Engine can be thought of as the autonomic data center of your complex distributed IT environment.


Resources

About the author

Lori Simcox is an Advisory Software Engineer in Tivoli Storage software development at IBM in San Jose, California. She received a B.S. degree in Mathematics from Pennsylvania State University and a M.S. degree in Computer Science from the University of Southern California. For the past eight years she has worked on IBM Tivoli Storage Manager and IBM TotalStorage Productivity Center. Her interests also include graphical user interface development and usability.

Comments



Trademarks

static.content.url=/developerworks/js/artrating/
SITE_ID=1
Zone=Tivoli
ArticleID=32119
ArticleTitle=Autonomic features of the IBM Virtualization Engine
publish-date=09142004
author1-email=lsimcox@us.ibm.com
author1-email-cc=