Performance Efficiency. Considering the function “list Files and Folders” and its definition “the function of returning the list of stored Files and Folders, with all the available details, if requested” provided in table 1 in section 3.3.1 in the following, an example of refinement of X sub characteristic measurement (Time Behaviour) is provided. In particular, time behaviour is meant as the degree to which the response and processing times and throughput rates of a product or system, when performing its functions, meet requirements. In the following, response time is adopted as possible measure for time behaviour and an example of application is provided. If response time is meant as duration from giving a command to start a batch of tasks till receiving the first response, then the Measurement function of X is: X = (B – A) /C A = time of entering a command to list stored files and folders B = time of receiving the first response C = time criteria specifying maximum allowable waiting duration from entering request to receiving response according to the software requirements X value smaller is better. Performance relative to the amount of resources used under stated conditions. Measurement function: PE= (X+Y+Z) / 3 where PE = Performance Efficiency X = Time behaviour Y = Resource utilization Z = Capacity
Performance Efficiency. Measurement function: PE = Z where PE = Performance Efficiency Z = Capacity Interpretation of test results: PE smaller is better Note: Useful element for the evaluation of the capacity can be: the number of requests or simultaneous access per unit of time; the number of simultaneous jobs accepted in the ingestion queue;r the number of tasks executed in parallel during a preservation workflow.
Performance Efficiency. This characteristic concerns the performances in terms of elapsed time, throughput and resource utilization for accomplishing specific tasks. This kind of tests make sense only if a reference system is fixed and considered. For LTFSArchiver the hardware/software system has been configured as follows: • Host HP Workstation Z800 o 8 cores 2.5 Ghz, o 8 GB RAM, o Linux release “Ubuntu 12.04 LTS” o LTFSArchiver 1.2 installed o AH401A PCIe 8Gb 2-Port Fibre Channel (QLogic) HBA for the connection to the HP LTO5 library and its drive o HP Smart Array P212 256M for the external SAS connection to the desktop LTO5 drives o Gigabit Ethernet connection to remote disk storage areas where the test files are stored • HP MSL2024 2 LTO-5 , LTO library with 24 slots and drive LTO5
Performance Efficiency. Since this is the first working prototype of the NOMAD platform, the performance efficiency is left for the next version of the NOMAD prototype and the results on this quality metric will be reported in D6.3, which is due M30. Specific log analysis screenshots will be also collected during the piloting and evaluation phase.
