The reports in this folder where conducted with the temporal-representation functionality of Orion-LD enabled. Only update tests where conducted, since retrieval is not impacted by this feature. Configuration information can be found in the config-folders and include the used configuration for Timescale-DB.
Environments:
- Tiny - 1 CPU / 6 GB RAM Orion-LD, 8 CPU / 16 GB RAM Mongo-DB, 4 CPU / 16 GB RAM Timescale-DB
- Small - 2 CPU / 12 GB RAM Orion-LD, 16 CPU / 32 GB RAM Mongo-DB, 8 CPU / 32 GB RAM Timescale-DB
- Mid - 8 CPU / 48 GB RAM Orion-LD, 30 CPU / 48 GB RAM Mongo-DB, 16 CPU / 64 GB Ram Timescale-DB
- Large - 16 CPU / 96 GB RAM Orion-LD, 60 CPU / 96 GB RAM Mongo-DB, 32 CPU / 128 GB Ram Timescale-DB
In order to get the required performance form mongo-db, we would recommend the usage of ssd's for storage. Since the
tests where run on a GCP-enviornment, we need to create such a storage class first. The
kubernetes-manifest for that can be found in the config. It can be applied to the
cluster via: kubectl apply -f <CONFIG_SIZE>/config/storage.yaml