May 23, 2022, author: Lubomír Bulej
We have released a bugfix update of the Renaissance benchmark suite, fixing two
race conditions discovered in the
Apart from the bug fixes, there have been no other changes to the benchmark
code or the underlying libraries. The changes in the affected benchmarks are
relatively minor, and the unaffected benchmarks are identical to Renaissance
0.14. Consequently, there should be no significant changes in the workloads.
Both race conditions resulted in exceptions visible in the benchmark output,
but only the exception in the
reactors benchmark caused the benchmark to fail
in a difficult-to-ignore deadlock. The exception in
finagle-chirper caused a
particular future computation to terminate prematurely, but that did not cause
the entire benchmark to fail. We apologize for the inconvenience.
Any comments and contributions are welcome.