Cycle 2024/3 test plans

docs/cycle_20243_testing_on_hifi.md

@@ -0,0 +1,52 @@
+# Cycle 2024/3 testing on HIFI
+
+## Basic verification of data pipeline for detector data
+
+Verify that:
+
+- Trace data is being received from all expected digitisers
+- Frame metadata is identical across digitisers and is correct
+- Event formation runs
+- Frame aggregator runs
+
+## Creation of NeXus files triggered by IBEX/SECI and containing correct metadata
+
+Verify that:
+
+- NeXus writer is triggered by a run start message from IBEX/SECI
+- NeXus writer is triggered by a run stop message from IBEX/SECI
+- Metadata sent via IBEX/SECI is stored in the NeXus file correctly (this could be interpreted as one of each metadata type or all of the metadata required for analysis in WIMDA)
+
+## Measurement of data rate and compute resource usage
+
+Measure:
+
+- Network data rate between digitiser switch and Kafka broker
+- Resource used on broker by:
+   - Redpanda
+   - Event formation
+   - Aggregator
+   - NeXus writer
+
+## Suitability of NeXus file for downstream filtering
+
+Verify that:
+
+- NeXus files contain experimental/detector data for the appropriate time period (i.e. they contain all of the data for the run and no more)
+- NeXus files contain metadata for the appropriate time period
+- It is possible to sensibly align detector and metadata in time
+- NeXus file contains sufficient metadata for filtering and histogram generation
+
+## Basic verification of filtering and histogram generation
+
+Verify that the tool:
+
+- Reads NeXus event files (written by NeXus writer)
+- Filter by frame timestamp
+- Filter by amplitude
+- Writes a NeXus file that WIMDA can correctly read
+
+## Verification/inspection of produced events and histograms
+
+Some form of manual inspection of the histograms, comparing them to those captured by the existing data acquisition.
+Details TBC.

docs/cycle_20243_testing_on_musr.md

@@ -0,0 +1,17 @@
+# Cycle 2024/3 testing on MuSR
+
+## Kubernetes based deployment of pipeline components
+
+Verify that:
+
+- Pipeline components can be created and function correctly once started (i.e. start, and process data from Kafka)
+- Relevant components (i.e. event formation) can be scaled and message processing functions as expected (i.e. messages are processed only once and distributed across all instances)
+- Pipeline components can be terminated without failure
+
+## Prometheus based monitoring
+
+Verify that:
+
+- Redpanda metrics are collected
+- Metrics of pipeline components are collected
+- Metrics are written correctly to InfluxDB via Telegraf