Details
-
Architectural Decision
-
Resolution: Done
-
None
Description
From SKA-TEL-SKO-0000008-12 SKA Phase 1 System requirements Specification:
SKA1-SYS_REQ-3003: SKA1_Mid shall support the designation of any subarray as an engineering subarray at Scheduling Block boundaries.
SKA1-SYS_REQ-3002: SKA1_Low shall support the designation of any subarray as an engineering subarray at Scheduling Block boundaries.
"Intended to be used for testing, commissioning, and maintenance. Resources marked anything other than operational may only be allocated to engineering subarrays."
"A subarray (also referred to as sub-array) is a subdivision of a SKA telescope that can be scheduled and be operated independently of other subarrays. A subarray constitutes a set of resources (i.e. receptors, correlator slices…) and can be as large as the whole telescope array, or a single constituent item. A subarray is only prevented from being created by resource constraints. Resources in this context include but are not limited to: number of available dishes/stations (receptors), power availability, data transport bandwidth, signal processing, or data processing capacity.
An Engineering subarray may not need an entire system slice, but this is not precluded. For instance, a set of dishes without the need of signal or data processing may constitute and Engineering subarray. Conversely, an Astronomy subarray requires end-to-end capability, i.e. a full system "slice"."
---
An Engineering subarray is a subarray that can be used for commissioning, maintenance, testing, including testing of new versions of H/W, F/W and/or S/W, while the rest of the Telescope is used for scientific observations.
There is a need to define detailed requirements for support of engineering subarrays, determine desired functionality and behavior , and implementation details.
Why now: For some of the major subsystems implementation of engineering subarray may be a challenge, and it is the best to define the requirement and identify limitations as early as possible to avoid costly re-work.
