Configuring AMI Data Export (ADE) system settings

  1. Go to System Administration > Sys Admin > System Settings.

  2. From the Section dropdown menu, select AMI Billing Export.

  3. Configure the following parameters, and then click Save.

AMI Data Export Dispatcher Settings:

  • Data Availability Re-check Frequency in Minutes. Defines the number minutes to wait before re-checking whether data is available to fulfill an export. This check is only performed after the initial data availability check if the initial check did not find data. Affects the latency and overall load on the database.

  • Entity Batching Delay in Seconds. Defines the number of seconds to wait for batches of requests to be prepared before writing them out to tasks. Affects the latency of requests flowing through the system. Also affects the number of tasks that are created.

  • Maximum Request Items Per Task. Defines the maximum number of request items to be processed by a task. Affects the number of tasks that get created. A smaller request size may result in more tasks.

  • Number of Configuration Fetch Threads. Defines the number of threads that are allocated to perform concurrent data availability fetches. This setting is typically dependent on the total number of processors available. Affects the overall throughput.

  • Number of Data Availability Fetch Threads. Defines the number of threads that are allocated to perform concurrent data availability fetches. This setting is typically dependent on the total number of processors available. Affects the overall throughput.

  • Number of Task Writer Threads. Defines the number of threads that are allocated to perform concurrent task creation. This setting is typically dependent on the total number of processors available. Affects the overall throughput.

  • Processing Chunk Size. Defines the number of request items grouped for internal processing in the ADE export service. Affects the number of entities that are retrieved from the AdeEntity database table during configuration fetches and data availability fetches. Affects the number of entities that are retrieved from the database during configuration fetches and data availability fetches.

    • Chunk size is a parameter common to several task templates. The higher the chunk size value, the greater the portion of the total available work each IEE task runner performs. A smaller value means dividing the work between more concurrent task runners.

    • Chunk sizes that are too large overburden a few task runners, while other task runners sit idle. In extreme cases, overly large chunk sizes could result in system failure due to lack of available, allocated memory. Chunk sizes that are too small spawn many task runners that perform more initialization and shutdown than actual work. Both large and small extremes will cause delays in export processing.

    • When fine tuning chunk sizes, consider database resources, hardware requirements, daily processing levels, and processing windows. Adjust values over time, as hardware and networking conditions change.

  • Request Polling Frequency in Seconds. Defines the number of seconds to wait between queries of the Export Dispatcher for requesting work items. Affects the latency of the initial request.

    • This value controls the export response time. The system waits this amount of seconds before requesting the initial work item. You can lower this value when performing demonstrations in which you want a more immediate response. A higher value causes longer wait times between the export dispatcher task and the new request, but a higher value can also help balance the load on a system that serves high demands.