Archived Newsletters
|
White papers:
|
Major New Capacity Planning Enhancements and More in TeamQuest's Q4 Software ReleaseTeamQuest's Q4 software release (officially known as Release 10.1 PF 20081001) was produced in record time and is packed with capabilities on TeamQuest customers' wish lists, including quick and easy cross-platform capacity planning.
Cross-platform capacity planning can help you prepare for the plunge. The Q4 release includes a number of powerful additions to TeamQuest capabilities:
Smoother, easier cross-platform capacity planningTeamQuest is best known for our capacity planning software product, TeamQuest Model. TeamQuest Model is predictive software that accounts for important non-linear behavior that is missed by simpler trend analysis capacity planning techniques. It can perform not only discrete event simulation, but, more importantly, analytic modeling, which is simpler and less time-consuming, but capable of producing very accurate results. TeamQuest Model allows you to quickly evaluate alternative configurations when you are:
The Q4 release of TeamQuest Model Release 10.1 streamlines the process of predicting cross-platform migrations. Say, for example, you are planning a server consolidation effort moving Linux workloads to run on a Unix platform or perhaps the opposite, consolidating Unix workloads to run on Linux. With the Q4 release of TeamQuest Model, predicting the performance of your consolidated workloads on alternative platforms is child's play. Before the Q4 release, TeamQuest Model could be used for such cross-platform capacity planning, taking into account changes in I/O and CPU configuration, but determining the relative power of different CPU/OS platforms was an exercise left to the user. Now for many popular platforms TeamQuest Model can quickly make cross-platform predictions without the need to research relative performance measures. The relative performance measures come with TeamQuest Model right out of the box, and the TeamQuest Model user interface makes it easy to incorporate them into its predictions. More intricate modeling of additional multithreaded processorsPast versions of TeamQuest Model have had the ability to predict the multithreaded performance of POWER5 and POWER6 processors. The Q4 release brings increased accuracy when predicting the multithreaded performance of UltraSPARC T1 and T2 and SPARC64 VI as well as Intel Xeon and Pentium 4, and Itanium processors.
 Multi-threaded Processor Architecture Predicting the performance of multithreaded processors is complicated by the fact that there are a wide variety of different hardware implementations of multithreading. Each implementation has innate differences in performance due to variations in internal architecture. One processor type may have a different number of processor cores, for example, and a different number of execution threads per core than another processor type. And different internal architectures means that the threads and cores of one processor type will interact with one another differently and have different performance characteristics than those of another processor type. TeamQuest Model's predictive capability for multithreaded processors works together with its cross-platform capacity planning functionality, allowing measurements from one multithreaded processor environment to be used as the baseline for predicting performance in another multithreaded processor environment. So you can evaluate options such as consolidating workloads from multithreaded Xeon environments to an UltraSPARC platform or from UltraSPARC to Xeon, ensuring that the configuration you eventually choose will satisfy service levels while meeting other requirements such as cost and power consumption limitations. Not only does TeamQuest Model take variations in processor architecture into account, it also considers virtualization and partitioning configuration when making predictions. For example, TeamQuest Model knows the effect of turning on simultaneous multithreading for a particular partition in AIX or capping a partition at two processors. And we are continually working to add even more support for popular virtualization capabilities. New data collection and analysis specifically for IBM's WPARs on AIXIBM recently introduced WPARs (Workload Partitions) with AIX 6.1. AIX WPARs provide some of the isolation afforded by other virtualized environments, without introducing the additional overhead required to run an operating system image in each partition. The Q4 release brings improved WPAR performance analysis capabilities to TeamQuest Performance Software. For example, information regarding WPAR configuration, such as entitled CPU and memory limitations can now be kept in TeamQuest's Capacity Management Database (CDB). There are also CPU utilization statistics available on a per-WPAR basis, and the containing WPAR is noted with each process' performance data in the CDB, making it possible to do detailed performance analysis on a WPAR basis. Expanded Sybase performance analysisThe Q4 release introduces a new version of TeamQuest's Sybase agent to Release 10.1 of TeamQuest Performance Software. The new agent is called the Sybase ASE Agent, corresponding with Sybase's current name for their database software. It supports Sybase ASE 12.5 and 15.0. Platforms supported include AIX, HP-UX (PA-RISC & Itanium), Linux (Red Hat & SuSE), Solaris (SPARC & x86), and Windows. TeamQuest's new Sybase ASE agent can provide up-to-the-minute status reports on your Sybase servers, revealing, for example, the SQL commands that are the top consumers of processor or memory resources, or what locks are currently active. The agent also knows a lot about your database's use of storage. For each database you can tell how much storage is used, allocated, and free. You can determine whether a database has reached capacity because the allocated space has been consumed or if the devices where the database resides have been filled. And for databases that are distributed across devices, you can see how much storage a database is consuming on each device. These are just a few examples of the new analysis made possible by the new Sybase ASE agent. They represent some powerful new additions, but arguably the best thing about the agent is that it is tightly integrated part of the overall TeamQuest Performance Software suite, supplying data to TeamQuest's federated capacity management database for sophisticated analysis using tools like TeamQuest IT Service Analyzer and Reporter. Capacity planning using an Oracle Enterprise CDBTeamQuest's native Capacity Management Database (CDB) is implemented using proprietary technology to provide an extremely efficient repository for distributed performance data. You can optionally create departmental or enterprise capacity management databases containing aggregated subsets of distributed performance data. These more centralized subsets can facilitate an efficient departmental or enterprise analysis of performance. The departmental or enterprise CDBs can be implemented with either TeamQuest's proprietary database engine, or using Oracle. (Using Oracle allows for open access using custom or third party tools.) And beginning with the Q4 release from TeamQuest, an Oracle CDB can provide the baseline data for capacity planning with TeamQuest Model. (Prior to the Q4 release, TeamQuest Model required that baseline data reside in a proprietary CDB.) Many productivity enhancementsA number of productivity-enhancing user interface improvements have been incorporated into TeamQuest's new flagship analysis and reporting tools, TeamQuest IT Service Analyzer and Reporter. See the bulleted list of improvements below for details. Many other improvements |
 |
TeamQuest IT Service Analyzer
|
 |
TeamQuest Manager
|
 |
TeamQuest Model
|
 |
TeamQuest IT Service Reporter
|
 |
TeamQuest View
Product DownloadTeamQuest Release 10.1 PF 20081001 product downloads are available to licensed customers with active maintenance via the customer support area. If you are not yet a customer, click the appropriate link under Connect with TeamQuest in the right margin of this page. |
