High Performance
Benchmarks and scalability metrics for Staex MCC in large-scale deployments.
Benchmark Setup#
The benchmark environment consisted of two cloud instances: one Staex node on cloud node A and 12,800 nodes on cloud node B. Each instance was equipped with an AMD EPYC 16-core processor and 32 GiB of RAM.
Scalability Results#
The benchmark achieved 800 nodes per processor core. The first instance (acting as the bootnode) consumed just 25 MiB of memory. The second instance, running all 12,800 nodes, used approximately 6 GiB total -- roughly 0.48 MiB per node. This demonstrates that MCC can scale to very large fleets on commodity hardware.
Bootstrap Performance#
New nodes join the network rapidly, with bootstrap time measured in seconds. This is critical for deployments with frequent power cycles or intermittent connectivity, such as solar-powered sensors, mobile assets, or devices in remote locations that experience regular connectivity interruptions.