Bull Performance Reports For Capacity Management

  • Uploaded by: Derek E Weeks
  • 0
  • 0
  • April 2020
  • PDF

This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA


Overview

Download & View Bull Performance Reports For Capacity Management as PDF for free.

More details

  • Words: 2,751
  • Pages: 12
Performance Management Outlining the advantages of virtualization

-1-

«

Today, virtualization is unavoidable for consolidation projects and to achieve a rapid

return on your IT investments. With innovating virtualization functions, Bull Escala servers, based on Power processors and the AIX® operating system allow you to consolidate multiple applications on a single physical server, while optimizing available resources. Its flexibility allows you to rapidly and dynamically implement new applications, while significantly reducing costs. Although the benefits are clear, the dynamic nature of these virtualization technologies is nevertheless often perceived as a complexity factor when it comes to administration and service quality analysis. Implementing the appropriate performance management tools demystifies this complexity.

-2-

The challenges of virtualized infrastructures Outlining the advantages of virtualization on Bull Escala, with performance management solutions that are adapted to your needs, and simple to implement

Although virtualization has definite benefits, its dynamic nature is often regarded as a complexity factor in terms of management by administrators, as well as by decision makers who are more accustomed to traditional infrastructures for which they are using tools that do not recognize new technologies. A director or an IT manager must first ensure global service to users, by aligning information resources with real needs, while also managing costs. To do this, he or she must have a clear general view of the information system, normally distributed and composed of heterogeneous physical and virtual environments. He or she must also anticipate potential performance degradations, adapt quickly to new needs, and be transparent to the users, as well as plan for future investments. For an administrator of virtualized Bull Escala systems, resource management is a major concern. The different performance analysis tools included with the AIX® operating system generate a very large amount of data recuperated at regular intervals in each partition. It must then be consolidated before it can be used. This can be very time consuming.

-3-

The administrator needs to have a general view of CPU and memory usage, as well as graphs and reports, showing relevant metrics in regards to virtualization technologies on Escala systems. Faced with these challenges, he or she needs tools that are capable of analyzing the various collected information, in order to establish clear and precise reports on the capacity and actual utilization of systems. To avoid failures, they must enable rather proactive than reactive management. This enables anticipation and good decision making, for instance when managing risks tied to consolidation project, or planning for future investments. This document describes the various criteria for each of the tools and shows the solutions recommended by Bull in order to benefit from the advantages of virtualization on Bull Escala, with performance management solutions that are adapted to your needs, and simple to implement. It also shows how Bull, through its Competence Centers, helps customers to optimize the performance of their infrastructure, and assists them in their consolidation and virtualization projects on Bull Escala servers.

Criteria of choice With adapted tools, IT directors and administrators can plan and implement virtualization projects with ease and take advantage of all of the benefits, while reinforcing performance quality

Taking into account virtualization Bull's experience has shown that classic performance management tools cannot properly manage virtual servers, nor can they take into account the specificities of the technologies that are in use. For virtual Bull Escala servers, metrics must be appropriate at the partition level, as well as at the shared processor pool level, and at the machine's global level. Centralized management of heterogeneous environments The majority of companies adopting virtualization technologies do so for distributed and heterogeneous environments, with several coexisting operating systems. Common management of these physical and virtual environments is therefore required to ensure their infrastructure is managed efficiently. Ease of installation, implementation and administration It must be possible to rapidly implement the tools on all servers without interrupting activity, and without the need for local parameterization. Furthermore, the collection and analysis of information must not have any impact on performance. Constant collection and historization of data Identifying failures, bottlenecks and peak workloads requires the constant collection of metrics, in order to easily define different workload profiles. It must also enable the analysis of data tendencies in a data repository, which is made up of all of the information collected in a given period.

-4-

Information accessible to everyone IT managers should have a global view of the health status of their infrastructure, being able to identify availability and saturation problems, as well as under or over-utilised servers, in order to make good decisions in terms of investment or development. To do this, they need to have reports and graphs at their disposal which are easy to read and interpret. These also allow for a certain transparency towards their users in regards to achieving commitments. Detailed and precise analysis for administrators Administrators must have access to graphs and activity reports for each physical server, as well as to detailed analyses of the behaviour of each partition, and shared processor pool.

Bull Performance Report: a performance management suite Bull Performance Report take advantage of virtualization benefits with Bull Escala servers

With Bull Performance Report, organizations now have access to a complete performance management solution for their entire information system, including Bull Escala virtualized environments. It enables IT managers and administrators to access all of the necessary elements to optimize information resource allocation, as well as to manage proactively by allowing them to anticipate problems and future needs. It results in increased user satisfaction and an optimal return on investment. Bull Performance Report is a complete suite of tools that is available in two complementary editions:

Enterprise Edition for the IS

Bull Performance Report Enterprise Edition provides IT management with a global and summarized view of the information system, accounting for distributed and heterogeneous environments, in particular those supported by virtualized platforms such as Bull Escala. Bull Performance Report Server Edition provides Bull Escala systems administrators with information and detailed reports on all of the Escala machines, their partitions and their shared processor pool.

Server Edition for Escala

Reports

Collectors Central Repository

Virtualized Escala servers

-5-

Bull Performance Report Enterprise Edition Bull Performance Report Enterprise Edition offers IT and operations managers summary reports on the status and behaviour of their information system. They can also better allocate resources according to needs. Perfectly adapted to virtualized server environments, in particular Bull Escala servers, Bull Performance Report Enterprise Edition enables IT managers to effectively drive their entire infrastructure and deliver optimal service to users while controlling costs.

Bull Performance Report Enterprise Edition architecture is based on: • Collectors or agents installed on each of the physical servers or partitions, continuously collecting several hundreds of data coming from the partitions or physical servers • A centralized data repository grouping information of the 30 added value metrics, with data historization over time • A user portal enable access to various charts and reports published in HTML pages, thanks to intuitive navigation with a macroscopic view to obtain, if necessary, more detailed analyses. Different types of reports are available, including GQR reports (Global Quality Report) and VSR reports (Virtualized Server Report). Taking into account virtualized Bull Escala servers Installing a super collector at the level of one of the partitions or a virtualized Bull Escala server enables Bull Performance Report Enterprise Edition to carefully control the resources of different virtualized partitions.

-6-

This super collector communicates with the Hardware Management Console (HMC) to obtain metrics out of which the ones relating to the shared processor pool. Optimized management of heterogeneous environments with normalized metrics Bull Performance Report Enterprise Edition recognizes heterogeneous server environments with a centralized and homogeneous view, regardless of the type or version of the operating system, in order to compare servers or server groups. Proactive management and better communication with users Bull Performance Report Enterprise Edition summary reports offer immediate access to essential information. IT managers have a global view of the status of their infrastructure at a glance, which enables them to objectively identify failures that could have repercussions on the quality of service provided to users, and resolve them with the recommended intervention priorities.

Reports Reports Data DataMining MiningTool Tool

AIX

Super-collectors, Super-collectors,Collectors Collectors Central Centralrepository repository Windows/NT/2000/2003

Sun Solaris

Powered by Systar

HP UX/SGI IRIX

Linux Redhat

db Thanks to data historization, these reports also make it possible to compare the development of the capacity consumed versus the total capacity, and to identify the potential changes in behaviour. This enables a quick reaction to step up to new demands, to plan for new investments or to prepare consolidation projects. Finally, these reports, easily read by non-technical staff, enable better communication and greater transparency to the users. Non-intrusive collection technology Low level metrics are continuously collected by the collectors then analyzed and certified locally for each partition or physical server, to produce statistics of use for each of the 30 added value metrics.

-7-

These are only transferred once per hour to the centralized data repository. This prevents network performance degradation due to continuous data transfers, as well as the need for a high capacity database to store significant volumes of data. Packaging optimized for virtualization on Escala The price of a Bull Performance Report Enterprise Edition license takes into account the server class on which it will be installed. Particularly attractive, it is delivered with virtualized Escala servers, and includes: • A certain number of collectors • A super collector • GQR and VSR reports, which enable monitoring and optimizing the performance quality and the capacity management of virtualized servers.

Every day, week and month, Bull Performance Report Enterprise Edition generates a health status report for servers and applications with the Identification of critical days, weeks or months and highlight areas at risk. Completely configurable, it allows, for a rapid overview, to configure the forecast by server, by application, by geography or by technological group or interval of times.

The VSR reports make it possible to monitor the configuration and the global activity of servers by view. The standard navigation principles allow the access to individual reports for each server and each partition.

Detailed Bull Escala virtualized server reports will show: • The configuration of each of the server's partitions • The global utilization of resources for the entire Escala server: CPU, memory, disks and I/O • The allocation of resources among the

CPU Use, hour by hour 05/31/2007

Total CPU used by Partitions (compared to Virtualized Physical server CPU capacity)

-8-

May, Typical day

Bull Performance Report Server Edition Bull Performance Report Server Edition is a reporting tool entirely developed by Bull to respond to the inherent need to manage virtual servers. It provides a precise and global view of all of the processor resources of Bull Escala servers, from both a quantitative and qualitative point of view.

Bull Performance Report Server Edition is based on a Client/Server architecture. The Clients represent the Bull Escala platforms that need to be monitored. A unique Agent is installed in one of the partitions of each Client. Every minute, it collects metrics in the various partitions, and then transfers all of the collected information to the Server once a day. These are then analyzed and summarized in order to provide daily, weekly and monthly statistics, reports and graphs related to the activity of the resource processors of the various different partitions with shared processors, processor pool and the I/O server (VIOS). These are accessible from any terminal with a Web browser. Processor capacity analysis The reports make it possible to monitor, not only on daily basis but also overtime, the allocation of processor resources. To facilitate the report analysis, it is possible to define warning thresholds, making it immediately visible if these are ever crossed. With Bull Performance Report Server Edition, the administrator can easily establish: • The activity profiles of the different partitions as well as those of the shared processor pool, • Consumption peaks and troughs • Abnormal behaviour • Trends.

-9-

Optimizing processor resources Adapted metrics on Bull Escala virtualized servers enable administrators to optimize processor resources. Bull Performance Report Server Edition analysis reports on the use of processor resources enable the administrator to verify if the various partitions are correctly configured, and to easily evaluate if certain Bull Escala servers are under or over-used. Decision making help Bull Performance Report Server Edition enables the administrator to react quickly faced with problems and to make betterinformed decisions, for example the creation of a new partition on the Escala server to respond to new needs, or the activation of additional processors (Power On Demand resources) to avoid saturation problems. The quality of service provided to users is therefore improved. Ease of implementation Bull Performance Report Server Edition is implemented very quickly in a large number of Escala servers, without interruption or local parameterization. A single agent is installed per partitioned machine. Controlled storage of historical data On the Server side, only the data collected during the week are saved. Every 8 days the base is cleaned in order to keep only a history of reports and statistics.

Every day, week and month, Bull Performance Report Server Edition generates a summary report to monitor the use of resources of each of the shared processor partitions, as well as the processor pool. The administrator can get answers to the following questions: • Is this the best configuration? • What is the activity profile of each partition and processor pool ? • What are the average utilization values over the course of the day, overnight, over the last 24 hours ? • What are the moving averages within a time frames? The summary reports are accessible with a web browser.

• Have the warning and critical thresholds crossed ? • Is there a partition that regularly saturates the CPU resource of the entire machine ? • What are the maximum CPU resources that an uncapped partition can consume ? • Do we have enough resources to implement a new partition ?

The warning threshold corresponds to use of 70% of the total available capacity of the processor. The critical threshold corresponds to use of 90% of the total available capacity of the processor.

- 10 -

From consulting to support, a complete range of professional services Thanks to its expertise in the management of highly available complex infrastructures, Bull offers a wide range of performance management services in consulting, training, installation and support related to performance management. Whatever the objective may be, Bull experts help customers in optimizing the performance of their current infrastructures as well as support them in their virtualization and consolidation projects. Performance analysis and capacity management Starting with a complete and detailed inventory of the infrastructure in place, and an analysis of the current and future needs, Bull consultants help customers to reach the required performance level, in either virtual or traditional environments: • load simulation to check if the infrastructure in place is able to support not only the expected load but also the potential peak loads , • identification and removal of performance inhibitors, • optimizations with adjustments of the parameters for various elements concerned (servers and operating system, network, storage, applications...) • monitoring of performance developments overtime with trend analysis.

- 11 -

This analysis, which is delivered as a technical report to the customer, is followed by recommendations to optimize the current resources usage or to plan for future investments. Sizing consolidation projects, cost control The under-utilization of physical or virtual servers involves unjustified material and software costs. To avoid this, Bull consultants, thanks to their expertise and know-how, help customers to rationalize their infrastructure and to optimize their return on investment. They offer a range of services from an audit of the existing infrastructure, including a feasibility study until the effective implementation. The specific needs of each customer are closely analyzed, which enables Bull consultants to commit to performance levels and high availability.

For more information… • Contact your commercial representative or Bull partner • Visit our website www.bull.com

B-Performance_va1

©Bull SAS – September 2007 - RCS Versailles B 642 058 739 - All trademarks in this document are the property of their respective owners. Bull reserves the right to modify this document at any time without notice. Certain offers or offer components described in this document may not be available locally. Please contact your local Bull representative to learn more about offers available in your country. This document is not a contractual agreement.

Bull – rue Jean Jaurès - 78340 Les Clayes-sous-Bois– France

Related Documents


More Documents from "Gaurav Kumar"