Stellar Supercluster (SSC) is package for automated integration testing of stellar-core. It works by running multiple containerized core nodes in self-contained simulated networks, and feeding them traffic (and/or invoking their internal load-generation testing subsystem). It is a second-generation tool, replacing the functionality of an older and now-retired package called Stellar Core commander (SCC).
SSC Has the following differences from SCC:
-
SCC used local processes and docker daemons. SSC uses Kubernetes for greater scalability, automation and co-tenancy among users. See doc/kubernetes.md for some notes on Kubernetes.
-
SCC was written in Ruby and was fairly slow, fragile and typo-prone. SSC is written in F# for greater compile-time error checking, performance and IDE support. See doc/fsharp.md for some notes on F#.
SSC has been driving day-to-day integration testing and simulation experiments at SDF since late 2019, and is capable of testing much larger and much more complex scenarios than SCC was, while being easier to maintain and more robust to errors.
See doc/getting-started.md for brief instructions on how to use it.
See doc/contributing.md. Support for uses outside of SDF will be provided on a best-effort basis.
See doc/missions.md for a list of supported tests.
See doc/measuring-transaction-throughput.md for instructions on load testing using supercluster.
See doc/theoretical-max-tps.md for a table of our theoretical max TPS results by release, as well as instructions on how to reproduce those numbers.