Performance
On an indicative basis, here are a few measurements to give an idea of the performance characteristics that can be expected from the cache.
These measurements were made on a Samsung Galaxy S8 (2017), using v1.0.0-beta.1 of the library.
This is based on benchmarks that can be found here.
Nested query: a fairly nested query involving lists and fragments (~6000 records / ~2 MiB of data).
Large list query: a simple query returning 10000 items in a flat list, each item with 2 fields (~10000 records / ~1.5 MiB of data).
Cache writes
Durations include normalizing the generated models and writing to the cache (JSON parsing is not included).
Cache | Query type | Duration |
|---|---|---|
Memory | Nested | ~800 ms |
Memory | Large list | ~800 ms |
SQL | Nested | ~1500 ms |
SQL | Large list | ~1800 ms |
Cache reads
Durations include reading from the cache, de-normalizing, instantiating and populating the generated models.
Cache | Query type | Duration |
|---|---|---|
Memory | Nested | ~530 ms |
Memory | Large list | ~490 ms |
SQL | Nested | ~830 ms |
SQL | Large list | ~850 ms |