Skip to content

Fix potential port conflicts. #30508

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 7 commits into from
Jan 18, 2021
Merged

Fix potential port conflicts. #30508

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
a266284
fix random port test=develop
gongweibao Jan 14, 2021
e110233
fix
gongweibao Jan 15, 2021
21ac46d
Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into…
gongweibao Jan 17, 2021
2bebdb4
cleanup test=develop
gongweibao Jan 17, 2021
7eb500c
cleanup test=develop
gongweibao Jan 17, 2021
3b1a166
cleanup test=develop
gongweibao Jan 17, 2021
350b698
fix port test=develop
gongweibao Jan 17, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 5 additions & 2 deletions python/paddle/fluid/tests/unittests/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -516,7 +516,6 @@ if(WITH_DISTRIBUTE)
py_test_modules(test_launch_coverage MODULES test_launch_coverage)
endif()

bash_test_modules(test_fleet_launch_ps START_BASH test_fleet_launch_ps.sh ENVS PADDLE_BINARY_DIR=${PADDLE_BINARY_DIR})
bash_test_modules(test_fleetrun START_BASH test_fleetrun.sh ENVS PADDLE_BINARY_DIR=${PADDLE_BINARY_DIR})
bash_test_modules(test_fleet_run_random_port START_BASH test_fleet_run_random_port.sh ENVS PADDLE_BINARY_DIR=${PADDLE_BINARY_DIR})
bash_test_modules(test_fleet_launch_async START_BASH test_fleet_launch_async.sh ENVS PADDLE_BINARY_DIR=${PADDLE_BINARY_DIR})
Expand All @@ -527,8 +526,12 @@ if(WITH_DISTRIBUTE)
set(dist_ut_port 20001)
foreach(TEST_OP ${DIST_TEST_OPS})
bash_test_modules(${TEST_OP} START_BASH dist_test.sh SERIAL LABELS "RUN_TYPE=EXCLUSIVE" ENVS "PADDLE_DIST_UT_PORT=${dist_ut_port}")
MATH(EXPR dist_ut_port "${dist_ut_port}+50")
MATH(EXPR dist_ut_port "${dist_ut_port}+40")
if(dist_ut_port GREATER_EQUAL 22998)
message(FATAL_ERROR "available ports have been exhausted:${dist_ut_port}")
endif()
endforeach(TEST_OP)
bash_test_modules(test_fleet_launch_ps START_BASH test_fleet_launch_ps.sh SERIAL LABELS "RUN_TYPE=EXCLUSIVE" ENVS "PADDLE_DIST_UT_PORT=${dist_ut_port}" PADDLE_BINARY_DIR=${PADDLE_BINARY_DIR} )
endif(NOT APPLE)
endif()

Expand Down
7 changes: 5 additions & 2 deletions python/paddle/fluid/tests/unittests/test_fleet_launch_ps.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,9 @@
set -e

function test_launch_ps(){
server_port_0=${PADDLE_DIST_UT_PORT}
server_port_1=$(( PADDLE_DIST_UT_PORT + 1 ))
echo "server_port_0:${server_port_0} server_port_1=${server_port_1}"
python -m paddle.distributed.fleet.launch --server_num=2 --worker_num=2 fleet_ps_training.py 2> ut.elog
if grep -q "server are killed" ut.elog; then
echo "test pserver launch succeed"
Expand All @@ -25,15 +28,15 @@ function test_launch_ps(){
exit -1
fi

python -m paddle.distributed.fleet.launch --servers="127.0.0.1:6780,127.0.0.1:6781" --workers="127.0.0.1:6782,127.0.0.1:6783" fleet_ps_training.py 2> ut.elog
python -m paddle.distributed.fleet.launch --servers="127.0.0.1:${server_port_0},127.0.0.1:${server_port_1}" --workers="127.0.0.1:6782,127.0.0.1:6783" fleet_ps_training.py 2> ut.elog
if grep -q "server are killed" ut.elog; then
echo "test pserver launch succeed"
else
echo "test pserver launch failed"
exit -1
fi

python -m paddle.distributed.fleet.launch --servers="127.0.0.1:6780,127.0.0.1:6781" --workers="127.0.0.1,127.0.0.1" fleet_ps_training.py 2> ut.elog
python -m paddle.distributed.fleet.launch --servers="127.0.0.1:${server_port_0},127.0.0.1:${server_port_1}" --workers="127.0.0.1,127.0.0.1" fleet_ps_training.py 2> ut.elog
if grep -q "server are killed" ut.elog; then
echo "test pserver launch succeed"
else
Expand Down