Slony-I has had (thus far) three test suites:
These were introduced as part of the original Slony-I distribution, and induced load via running pgbench.
Unfortunately, the tests required human intervention to control invocation and shutdown of tests, so running them could not be readily automated.
Slony-I version 1.1.5, introduced a test framework intended to better support automation of the tests. It eliminated the use of xterm, and tests were self-contained and self-controlled, so that one could run a series of tests.
Unfortunately, the framework did not include any way of inducing distributed load, so as to test scenarios involving sophisticated concurrent activity.
Introduced during testing of Slony-I version 2.0 during 2010, and released in early 2011, this framework is intended to be a better replacement for all of the preceding test frameworks.