diff options
| author | tv <tv@krebsco.de> | 2026-11-04 21:49:15 +0100 |
|---|---|---|
| committer | tv <tv@krebsco.de> | 2026-11-04 21:49:15 +0100 |
| commit | 27c870ed85deb5b5a1c60e0d1ed831d0b1108d6b (patch) | |
| tree | 0f9e0666fcba4f6fc4a87a4d477990718871783a /docs/development.md | |
| parent | 0536a8260e7d0bc4f050b310b8bb4902da16fa59 (diff) | |
benchmark members joining
Diffstat (limited to 'docs/development.md')
| -rw-r--r-- | docs/development.md | 6 |
1 files changed, 3 insertions, 3 deletions
diff --git a/docs/development.md b/docs/development.md index db89c04..9107014 100644 --- a/docs/development.md +++ b/docs/development.md @@ -21,9 +21,9 @@ The benchmarks are ignored tests named `bench_*`: the document's (writes and syn messages, `document/bench.rs`), the remakes' replay (`remake/bench.rs`), and what passing changes on costs over the simulated network (`simtest/bench.rs`: messages and bytes per write by kind, propagation delays, for a full mesh, a writer behind NAT, two -transports, partial meshes of 50 and 100 nodes, lossy links, and a member away for a -while; several minutes, or one scenario -alone with `BENCH_SCENARIO`, part of its name: `BENCH_SCENARIO=NAT`; all at once with +transports, partial meshes of 50 and 100 nodes, lossy links, a member away for a +while, and members joining during the writes; several minutes, or one scenario alone +with `BENCH_SCENARIO`, part of its name: `BENCH_SCENARIO=NAT`; all at once with `BENCH_PARALLEL=1`, faster on many cores, though sharing them makes delays come out somewhat higher), and the same estimated on the member list's graph alone for up to 2000 members, with some |
