Skip to content

[DO NOT MERGE] RDMA CM FIXES #171

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

Open
quasiben wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[DO NOT MERGE] RDMA CM FIXES #171

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
27 changes: 18 additions & 9 deletions tpcx_bb/benchmark_runner/slurm/Dockerfile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,7 +62,6 @@ RUN set -x \
deb http://archive.ubuntu.com/ubuntu/ xenial universe\n\
deb http://archive.ubuntu.com/ubuntu/ xenial-updates universe\
'" >> /etc/apt/sources.list.d/xenial.list \
&& apt update -y || true && apt install -y libibcm-dev \
&& rm /etc/apt/sources.list.d/xenial.list \
# cleanup
&& rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
Expand All @@ -77,7 +76,8 @@ COPY /MLNX_OFED_LINUX-5.1-2.5.8.0-ubuntu18.04-x86_64.tgz /tmp/MLNX_OFED_LINUX-5.
RUN tar -xzf /tmp/MLNX_OFED_LINUX-5.1-2.5.8.0-ubuntu18.04-x86_64.tgz && \
cd MLNX_OFED_LINUX-5.1-2.5.8.0-ubuntu18.04-x86_64 && \
apt-get update && apt-get install -y --no-install-recommends \
./DEBS/libibverbs* ./DEBS/ibverbs-providers* \
./DEBS/libibverbs* ./DEBS/ibverbs-providers* ./DEBS/librdma* \
./DEBS/perftest* \
&& rm -rf /var/lib/apt/lists/* \
&& rm -rf /tmp/MLNX_OFED_LINUX*

Expand All @@ -89,14 +89,13 @@ RUN /opt/conda/bin/conda install -n $CONDA_ENV -c conda-forge autoconf cython au


# Install UCX
ADD https://raw.githubusercontent.com/rapidsai/ucx-split-feedstock/master/recipe/add-page-alignment.patch /tmp/add-page-alignment.patch
ADD https://raw.githubusercontent.com/rapidsai/ucx-split-feedstock/master/recipe/ib_registration_cache.patch /tmp/ib_registration_cache.patch
ADD https://raw.githubusercontent.com/rapidsai/ucx-split-feedstock/bd0377fb7363fd0ddbc3d506ae3414ef6f2e2f50/recipe/add-page-alignment.patch /tmp/add-page-alignment.patch
ADD https://raw.githubusercontent.com/rapidsai/ucx-split-feedstock/bd0377fb7363fd0ddbc3d506ae3414ef6f2e2f50/recipe/ib_registration_cache.patch /tmp/ib_registration_cache.patch

RUN git clone --recurse-submodules -b v1.8.x https://github.com/openucx/ucx /tmp/ucx \

RUN git clone --recurse-submodules -b v1.9.x https://github.com/openucx/ucx /tmp/ucx \
&& cd /tmp/ucx \
&& source activate $CONDA_ENV \
&& patch -p1 < /tmp/ib_registration_cache.patch \
&& patch -p1 < /tmp/add-page-alignment.patch \
&& ./autogen.sh \
&& ./configure \
--prefix="${CONDA_PREFIX}" \
Expand All @@ -105,7 +104,6 @@ RUN git clone --recurse-submodules -b v1.8.x https://github.com/openucx/ucx /tmp
--enable-mt \
--enable-numa \
--with-gnu-ld \
--with-cm \
--with-rdmacm \
--with-verbs \
--with-cuda="${CUDA_HOME}" \
Expand All @@ -114,13 +112,24 @@ RUN git clone --recurse-submodules -b v1.8.x https://github.com/openucx/ucx /tmp

# Install UCX-py
RUN set -x \
&& git clone --recurse-submodules -j${PARALLEL_LEVEL} -b branch-0.16 https://github.com/rapidsai/ucx-py.git /tmp/ucx-py \
&& git clone --recurse-submodules -j${PARALLEL_LEVEL} -b branch-0.17 https://github.com/rapidsai/ucx-py.git /tmp/ucx-py \
&& source activate $CONDA_ENV \
&& cd /tmp/ucx-py \
&& python setup.py build_ext --inplace \
&& python -m pip install . --no-deps --no-cache-dir -vvvv \
&& rm -rf /tmp/ucx-py


## Install Latest Dask/Distributed
RUN set -x \
&& source activate $CONDA_ENV \
&& git clone https://github.com/dask/dask.git /tmp/dask \
&& git clone https://github.com/dask/distributed.git /tmp/distributed \
&& cd /tmp/dask && python -m pip install . \
&& cd /tmp/distributed \
&& python -m pip install -vv --no-deps --install-option="--with-cython=profile" . \
&& rm -rf /tmp/dask /tmp/distributed

# use login shell when exec'ing into the container
# SHELL ["/bin/bash", "-l"]

Expand Down