Previously we were calling bulkUpdate on all participants regardless of whether the fetched participants were different from the ones stored in IndexedDB. Since the participant data changes infrequently, we were wasting a lot of CPU cycles on updating identical records.
We now filter out only those participants that are actually different and only bulkUpdate those.
github-actions[bot]
commented
3 months ago
Previously we were calling
bulkUpdateon all participants regardless of whether the fetched participants were different from the ones stored in IndexedDB. Since the participant data changes infrequently, we were wasting a lot of CPU cycles on updating identical records.We now filter out only those participants that are actually different and only
bulkUpdatethose.For 10k participants, this leads to ~6x speedup.