Design considerations

Before installing Edge Data Store (EDS (Edge Data Store)), determine your storage and throughput needs and select devices that meet those needs.

EDS Storage role

The EDS Storage component is integrated with the EDS (Edge Data Store) and does not replace any existing storage technology produced by AVEVA. The EDS Storage component is a resilient and reliable data store, but is limited in the duration and scope of the data it retains.

• By default, the storage component processes data in a FIFO (first in, first out) (first in first out) method: as new data comes in and the size of streams exceeds the configured limits, older data is purged.

• You must egress data that you need permanently retained to either PI Data Archive, using the PI Web API Open Message Format (OMF (Open Message Format)) endpoint, or to CONNECT data services, using the CONNECT data services OMF (Open Message Format) ingress endpoint.

EDS Storage scale

The EDS Storage component provides an appropriate level of storage performance for small devices.

• For the smallest of these devices, throughput may be limited to tens of events per second.

• For larger devices with faster processors, memory, and storage, this could increase up to 3,000 events per second. The EDS Storage component is designed for small devices in Edge scenarios. If high throughput or large stream counts are required, CONNECT data services or PI Data Archive are more appropriate choices.

  Edge

Sizing of Edge devices

For EDS (Edge Data Store), there are three supported tiers of performance:

• Small Devices: 1 Core CPU (Central Processing Unit), 512 MB RAM (Random-Access Memory). 30 events/second, 200 streams total.

• Medium Devices: 2 Core CPU (Central Processing Unit), 1 GB RAM (Random-Access Memory). 300 events/second, 2000 streams total.

• Large Devices: 4 Core CPU (Central Processing Unit), 4 GB RAM (Random-Access Memory), SSD (Solid State Drive) storage. 3000 events/second, 3000 streams total.

These performance metrics assume solid state storage, which is commonly used in Edge devices.