Case Study: SkyElectric achieves millisecond performance and operational simplicity with ScyllaDB

SkyElectric Uses Scylla to Power Its Smart Energy Platform

SkyElectric, a smart solar energy provider, was running legacy Java applications on MySQL and struggling with rapidly growing IoT time‑series data that caused queries to take up to five minutes. While evaluating time‑series alternatives, the team rejected Riak TS and, after testing Cassandra, chose ScyllaDB’s Scylla database for its C++ implementation and promise of higher performance and operational simplicity.

SkyElectric deployed a 3‑node Scylla cluster on AWS (moving to larger i3.large instances) and used ScyllaDB support via Slack; the migration required little tuning and proved operationally effortless. The results were dramatic: roughly 10x better performance, average write latency reduced to 1.4 ms (from ~2 s), average read latency under 1 ms (previously often minutes), 5x higher request throughput, and 3x storage with no performance degradation, enabling the team to focus on new features with confidence in ScyllaDB’s reliability.

SkyElectric

Meraj Rasool

DevOps Lead