From dc56e21d5114e7b05c141feecc36fefe88ebc5bf Mon Sep 17 00:00:00 2001 From: cpattigi Date: Wed, 6 Aug 2025 11:19:14 +0000 Subject: [PATCH 1/2] added retry --- .../tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 | 2 +- .../tools/tf_sig_build_dockerfiles/setup.packages.rocm.el8.sh | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 b/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 index 86fa30dfaf22f2..ac0dbd3d93be73 100644 --- a/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 +++ b/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 @@ -3,7 +3,7 @@ FROM quay.io/pypa/manylinux_2_28_x86_64:2023-11-29-1ba608e as builder ################################################################################ # Add epel and elrepo repos and enable powertools repo -RUN dnf update -y && dnf install -y epel-release && dnf install -y elrepo-release && dnf config-manager --set-enabled powertools +RUN retries=3; for i in $(seq 1 $retries); do dnf update -y && break || sleep $((i * 5)); done && dnf install -y epel-release && dnf install -y elrepo-release && dnf config-manager --set-enabled powertools # Setup the build_system repo RUN echo -e "[build_system]\nname=ROCm\nbaseurl=https://repo.almalinux.org/build_system/8/x86_64/\nenabled=1\ngpgcheck=0" >/etc/yum.repos.d/build_system.repo diff --git a/tensorflow/tools/tf_sig_build_dockerfiles/setup.packages.rocm.el8.sh b/tensorflow/tools/tf_sig_build_dockerfiles/setup.packages.rocm.el8.sh index 00386b7653449b..4daf2460c4f0e1 100755 --- a/tensorflow/tools/tf_sig_build_dockerfiles/setup.packages.rocm.el8.sh +++ b/tensorflow/tools/tf_sig_build_dockerfiles/setup.packages.rocm.el8.sh @@ -21,6 +21,6 @@ set -e set -x dnf clean all -dnf update -y +retries=3; for i in $(seq 1 $retries); do dnf update -y && break || sleep $((i * 5)); done PKG_LIST=$(sed -e '/^\s*#.*$/d' -e '/^\s*$/d' "$1" | sort -u) dnf --enablerepo=extras,epel,elrepo,powertools,build_system install -y $PKG_LIST From 8648c61553486ead45d25c01337a79c57ab995c2 Mon Sep 17 00:00:00 2001 From: cpattigi Date: Wed, 6 Aug 2025 11:54:41 +0000 Subject: [PATCH 2/2] add more rtry --- .../tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 b/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 index ac0dbd3d93be73..31540715e16a89 100644 --- a/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 +++ b/tensorflow/tools/tf_sig_build_dockerfiles/Dockerfile.rocm.manylinux_2_28 @@ -10,7 +10,7 @@ RUN echo -e "[build_system]\nname=ROCm\nbaseurl=https://repo.almalinux.org/build # Setup c++ dev # TODO: Only install what is needed -RUN dnf group install -y "Development Tools" && dnf install -y llvm-toolset gcc-toolset-13 +RUN retries=3; for i in $(seq 1 $retries); do dnf group install -y "Development Tools" && break || sleep $((i * 5)); done && dnf install -y llvm-toolset gcc-toolset-13 # Install dependencies COPY setup.packages.rocm.el8.sh setup.packages.rocm.el8.sh