The data models should be broken out by locale. With multiple locales stored in S3, we will have substantial data. Hundreds of MB, possible a GB. We need to put a redis cache between taar-lite and the s3 dataset to improve data load times after the initial data load.
The data models should be broken out by locale. With multiple locales stored in S3, we will have substantial data. Hundreds of MB, possible a GB. We need to put a redis cache between taar-lite and the s3 dataset to improve data load times after the initial data load.