How do these performance attributes translate into tangible business benefits? While the document is tailored to IBM i clients, much of the introductory material in the performance section is OS-agnostic. While response time and throughput are generally related—increasing throughput normally leads to a decrease in response time—a system can throw a lot of resource at some amount of processing, causing a decrease in response time at the expense of throughput to other processing. The IBM Power Systems platform is built with a balanced system design to maintain good throughput and responsiveness—that is, good performance. With this understanding of performance as being driven by a mix of variables, we can start to make comparisons.
|Published (Last):||21 October 2005|
|PDF File Size:||1.97 Mb|
|ePub File Size:||16.46 Mb|
|Price:||Free* [*Free Regsitration Required]|
The Generic Workload can be a very versatile workload. It provides a controlled manner in which the user is allowed to enter specific resource requirements for characterizing a workload. These parameters include such things as rPerf and CPW processing requirements of a workload, the amount of memory, the number of disk arms, the amount of disk storage needed by the workload.
Back to top. The processor section refers to the resource requirements for the central processor. This requirement is specified differently depending on the operating system of the partition. This is the amount of memory in MB that is needed for this work to perform acceptably. The generic workload allows you to enter a base disk configuration and have it converted to a disk configuration with different characteristics, if you select "Yes".
In this case, configuration refers to attachment type i. If you enter known disk requirements with a known disk configuration, Workload Estimator can convert these to configuration characteristics that you would like to size. When choosing to convert the given Disk Configuration to match the disk options specified in User Options, the configuration entered will be modified before being considered in the solution selection process. The default value and the general recommendation for this is "Yes".
However, if the configuration being specified is intended to be added to the Selected System without any conversion, select "No". Be careful when doing this for a number of reasons. That particular level of hardware might be outdated which might limit future purchases. The Disk Configuration section describes the known characteristics of your disks for your workload that you are sizing, such as the type of disk drives, the style of storage protection used, the number the drives, the disk busy percentage, the amount of storage used, and the number and size of the disk operations.
As you describe your workloads, it may be appropriate to separate your disks into groups e. Up to eight separate disk groupings can be used. Information concerning the use of the screen control that is used to enter this information can be found in the Disk Group Applet tutorial. The Estimator validates the amount of drives that you specify against the maximum number of disk drives for the processor model that you specified.
For the workload that you are describing, please enter the following information that describe its characteristics:. Several years ago the WLE disk sizing support was enhanced by adding the Disk operation method ops method. When doing so, it is important to have that additional detailed information to establish ops rate, R:W ratio, and data rate; this helps consider the effectiveness of the cache, and other unique performance characteristics of each disk attachment type.
For the last couple of years, we have left in the support for the arms method input of drives and utilization for a bridge. Considering the above rationale, the arms method is only appropriate if mapping back to the same type of disk attachment; it provides more risk when mapping to other attachment types.
The GUI has said that the ops method is preferred. With WLE version We hope that you can continue to use WLE for sizing using the preferred approach of the ops method. For many of the built-in workloads, journaling and any associated resource consumption is already integrated into the workload. In cases where you have measured data without journaling support and you would like to add journaling support, please use the existing or generic workloads.
In this case, the Estimator uses a very high-level technique for increasing resources to represent journaling. The additional processor, memory, and disk resources required to support journal transactions are estimated based on the number of disk writes from the specific workload. Actual results may vary significantly as there is a wide range of potential settings. The Estimator assumes that all data accesses are considered for journaling and that reasonable performance tuning is done disk configuration is adequate, typical use of journal cache, etc.
There is other documentation available to help understand the performance implications of journaling with commitment control; please seek addition assistance if you want to better understand your specific implementation. If you would like to add journal with commitment control to your generic or existing workloads, click yes.
If your base measurements already include journaling, then you should click no, as the resources are already being considered. Keep in mind that this is a very course representation of this function as it adds resources based on the observed disk write values.
Skip to main content. The Generic Workload can be used in these cases: There are other analysis or sizing tools that have provided specific resource requirements i. Now you can incorporate this information into the Estimator so a complete sizing can be continued.
You would like to include additional resources that will be accounted for in the estimation. Examples of this would be to include more disk storage, or memory, etc.
You would like to use the disk configuration conversion functions provided by this workload. For example, you may wish to see what the requirements are for converting from one method of storage protection to another.
Definitions Processor The processor section refers to the resource requirements for the central processor. Non-Interactive CPW -- The non-interactive CPW represents the relative processor performance required to perform the non-interactive portion of the work.
Because there is not a Linux relative processor rating, WLE will allow you to enter your processor requirements using rPerf. Regardless, please do so carefully; this is a very rough mapping with no specific claim to actual performance or scaling. The rPerf model is not intended to represent any specific public benchmark results and should not be reasonably used in that way.
Memory This is the amount of memory in MB that is needed for this work to perform acceptably. Converting Disk Configuration The generic workload allows you to enter a base disk configuration and have it converted to a disk configuration with different characteristics, if you select "Yes". Disk Configuration The Disk Configuration section describes the known characteristics of your disks for your workload that you are sizing, such as the type of disk drives, the style of storage protection used, the number the drives, the disk busy percentage, the amount of storage used, and the number and size of the disk operations.
For the workload that you are describing, please enter the following information that describe its characteristics: Disk Group Name: Provide a meaningful name to represent this disk group. Disk Attachment Type: Select the disk attachment method. Disk Drive Speed: Select the disk speed. If mirrored is selected, the Estimator is assuming hardware mirroring, except software mirroring for IBM i.
This assumption is important for calculations for operations per second. Storage GB : Select the total amount of storage consumed not total space available. Make sure this amount matches the storage protection type as input. To reflect disk activity, provide the following input: Disk operation method: Read Ops Per Second: Enter the total number of disk read operations per second. Typically, this method of counting disk ops matches the information given by performance tools from the operating system.
Bytes Per Read Op: Enter the average size of these disk read operations. If uncertain, enter a typical value like Note that this parameter is only used if external is selected for attachment type. Write Ops Per Second: Enter the total number of disk write operations per second.
Reap the Performance Benefits of Upgrading from POWER7 to POWER9
Actual performance will vary based on application and configuration details. Although rPerf may be used to compare estimated IBM UNIX commercial processing performance, actual system performance may vary and is dependent upon many factors including system hardware configuration and software design and configuration. Variations in incremental system performance may be observed in commercial workloads due to changes in the underlying system architecture. Sometimes you may want to compare the rPerf of different IBM Power Systems, some of which you may not even have access to. If you're replacing a POWER6 system with POWER8, you'll need some way to compare these systems so, for instance, you'll have a better understanding of the number of cores you'll need to activate for new or existing workloads. As long as you know the model number, the total number of cores on the system and the CPU speed, you can obtain some valuable information.
More rPerf Resources
Check here to start a new keyword search. Search support or find a product: Search. Search results are not available at this time. Please try again later or use one of the other support options on this page. Watson Product Search Search. None of the above, continue with my search.
IBM Workload Estimator 2019.1