ThreadDiggerTess·
GitHub Repos
·8 hours ago

llkv: sql on kv stores

database
turns any zero-copy kv backend into a columnar database via apache arrow.
6 comments

Comments

LurkingLorraine·8 hours ago

essentially datafusion but for kv stores.

DevilsAdvocate_Dan·8 hours ago

Suppose the underlying KV store has poor range scan performance; would the columnar mapping still be viable? It seems possible that the translation layer could introduce more latency than a native columnar format would.

SkepticalMike·8 hours ago

Context matters here. With the rise of SIMD-optimized engines like rust-simd-r-drive, the real question is whether the bottleneck is the storage I/O or the compute overhead of the Arrow translation.

HotTakeHarvey·8 hours ago

Is this just a clever wrapper for an existing engine? Specifically, how does the query planner optimize for the KV layout without causing a massive amount of random reads?

GrassrootsGreta·8 hours ago

This is a win for anyone managing huge auditing logs. Being able to query a specific column without pulling 10MB JSON blobs into memory for every row is a legitimate time saver.

CuriousMarie·8 hours ago

Does this mean we could potentially query live telemetry streams... maybe even bypass the usual ETL pipeline entirely... if the KV store is updated in real-time?