HiveCacheStore + dio_cache_interceptor — the 3 traps nobody documents
Cold-start corruption, no TTL purge on launch, unbounded size — three edge cases of Dio + Hive caching in Flutter to handle from day one to avoid breakage.
HiveCacheStore + dio_cache_interceptor is convenient. But there are 3 traps nobody documents.
⚠️ Cold-start corruption
If the Hive box is corrupted (app update, schema change), it crashes silently. Always wrap the init in a try/catch with box deletion as a fallback.
⚠️ No TTL purge at startup
Expired entries are not cleaned on open. They're only cleaned on the next request for the same key. A store.clean(staleOnly: true) at startup keeps the cache from growing needlessly.
⚠️ Unbounded size
Hive doesn't limit the box size. On an app with many endpoints, it can eat storage without you noticing.
Bottom line
Dio + Hive is solid. But only if you handle these edge cases from day one.