Optimizing JSON Storage in Redis for Feed-Based Systems

Hey,

Hope you're keeping well.

For feed-based systems in .NET, storing large JSON blobs in Redis as plain strings is rarely optimal. If you need partial updates or selective retrieval, use RedisJSON with NRedisStack so you can update paths atomically without reserializing the entire object. For memory efficiency, serialize with System.Text.Json using JsonSerializerOptions to trim unnecessary properties, and consider compressing with Zstd via ZstdSharp before storing. If feeds are immutable once written, splitting them into smaller keyed chunks (e.g., per user or time window) can reduce eviction impact and improve lookup speed under allkeys-lru or volatile-ttl. Always benchmark serialization, compression, and Redis command latency in your ASP.NET Core application before committing to a storage strategy.

Thanks and regards,

Taz

1 Reply

Know the answer? Post it — somebody with the same question will find it here.

Tasadduq Burney9mo ago
Hey,
Hope you're keeping well.
For feed-based systems in .NET, storing large JSON blobs in Redis as plain strings is rarely optimal. If you need partial updates or selective retrieval, use RedisJSON with NRedisStack so you can update paths atomically without reserializing the entire object. For memory efficiency, serialize with System.Text.Json using JsonSerializerOptions to trim unnecessary properties, and consider compressing with Zstd via ZstdSharp before storing. If feeds are immutable once written, splitting them into smaller keyed chunks (e.g., per user or time window) can reduce eviction impact and improve lookup speed under allkeys-lru or volatile-ttl. Always benchmark serialization, compression, and Redis command latency in your ASP.NET Core application before committing to a storage strategy.
Thanks and regards,
Taz