diff options
| author | Garrett D'Amore <garrett@damore.org> | 2020-12-26 11:54:40 -0800 |
|---|---|---|
| committer | Garrett D'Amore <garrett@damore.org> | 2020-12-27 17:26:19 -0800 |
| commit | 3d535b638667ad0fcfff4246fce61c0176a056c4 (patch) | |
| tree | ec9b5054885e3898ac0f5f3ed203ab49534d11bb /perf | |
| parent | c3bfec2b38caaf34379a891e0ea30c7e48147c6f (diff) | |
| download | nng-3d535b638667ad0fcfff4246fce61c0176a056c4.tar.gz nng-3d535b638667ad0fcfff4246fce61c0176a056c4.tar.bz2 nng-3d535b638667ad0fcfff4246fce61c0176a056c4.zip | |
fixes #972 Very slow pull/push performance compared to ZMQ
This refactors the pipeline protocol to use lightweight mq
instead of the more expensive message queue structure. It
also provides nicer backpressure and buffering support.
The test suite was updated and converted to NUTS as well.
This won't completely close the gap, but it should help quite
a bit.
Diffstat (limited to 'perf')
| -rw-r--r-- | perf/perf.c | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/perf/perf.c b/perf/perf.c index f1995aba..accac621 100644 --- a/perf/perf.c +++ b/perf/perf.c @@ -419,6 +419,10 @@ do_inproc_thr(int argc, char **argv) open_client = nng_pair1_open; open_server = nng_pair1_open; break; + case OPT_PIPELINE0: + open_client = nng_pull0_open; + open_server = nng_push0_open; + break; case OPT_URL: addr = arg; break; |
