Processor Cores and Licenses
The licensing of Altova server products is based on the number of physical processor cores available on the product machine (as opposed to the number of logical cores). For example, a dual-core processor has two cores, a quad-core processor four cores, a hexa-core processor six cores, and so on. The number of cores licensed for a product must be greater than or equal to the number of cores available on that server machine, whether the server is a physical or virtual machine. For example, if a server has eight cores, you must purchase an 8-core license. You can also combine licenses to achieve the core count. So, two 4-core licenses can be used for an eight-core server instead of one 8-core license.
If you are using a computer server with a large number of CPU cores but only have a low volume to process, you may also create a virtual machine that is allocated a smaller number of cores, and purchase a license for that number. Such a deployment, of course, will have less processing speed than if all available cores of the computer were utilized.
|Note:||Each Altova server product license can be used for only one client machine at a time—the machine on which the Altova server product is installed—even if the license has unused licensing capacity. For example, if a 10-core license is used for a client machine that has 6 CPU cores, then the remaining 4 cores of licensing capacity cannot be used simultaneously for another client machine.|
Note for MobileTogether Server Advanced Edition
Because of its services functionality, MobileTogether Server Advanced Edition will run only on machines with two or more cores.
Note for FlowForce Server and MapForce Server
FlowForce Server Advanced Edition and MapForce Server Advanced Edition will run only on machines with two or more cores.
When assessing the number of cores you should license, this decision should take into account the data volumes you need to process and the processing time your business environment is expected to allow for. In most scenarios, a larger number of cores means larger amounts of data processed in less time. Below are just a few application-specific tips:
•FlowForce Server runs as a multi-threaded application. If the number of concurrent requests to the server is large, an insufficient number of cores will lead to latency (waiting times). For example, if you are exposing jobs as Web services, there may be hundreds of concurrent requests from clients. In this case, FlowForce Server will significantly benefit from a larger number of cores.
•MapForce Server will utilize a single core at a time, per mapping. Therefore, if you need to run multiple mappings simultaneously, a larger number of cores is highly recommended. For example, when MapForce Server runs under FlowForce Server management, several mapping jobs may overlap and run concurrently, depending also on the setup. Note, however, that if the volume of data processed by your mappings is extremely large, then latency could still occur.
If an Altova server product allows single-thread execution, an option for Single-thread execution will be available. In these cases, if an Altova server-product license for only one core is available in the license pool, a machine with multiple cores can be assigned this one-core license. In such a case, the machine will run that product on a single core. Processing will therefore be slower, because multi-threading (which is possible on multiple cores) will not be available. The product will be executed in single thread mode on that machine.
To assign a single-core license to a multiple-core machine, in LicenseServer, select the Limit to single thread execution check box for that product.
Estimate of core requirements
There are various external factors that influence the data volumes and processing times your server can handle (for example: the hardware, the current load on the CPU, and memory allocation of other applications running on the server). In order to measure performance as accurately as possible, test the applications in your environment with data volumes and in conditions that approximate as closely as possible to real business situations.