mirror of
https://github.com/checkpoint-restore/criu.git
synced 2026-01-23 02:14:37 +00:00
30acbabcdd
Docker version 28 broke container restore in combination with network namespaces. The workaround in the CI script was excluding Docker version 28. Now that there is also Docker version 29, which is still broken, this also excludes Docker version 29. Signed-off-by: Adrian Reber <areber@redhat.com>
139 lines
3.6 KiB
Bash
Executable file
139 lines
3.6 KiB
Bash
Executable file
#!/bin/bash
|
|
|
|
set -x -e -o pipefail
|
|
|
|
# Workaround: Docker 28.x and 29.x has a known regression that breaks the checkpoint and
|
|
# restore (C/R) feature. Let's install previous, or next major version. See
|
|
# https://github.com/moby/moby/issues/50750 for details on the bug.
|
|
export DEBIAN_FRONTEND=noninteractive
|
|
apt remove -y docker-ce docker-ce-cli
|
|
../../contrib/apt-install -y ca-certificates curl
|
|
install -m 0755 -d /etc/apt/keyrings
|
|
curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.asc
|
|
chmod a+r /etc/apt/keyrings/docker.asc
|
|
# shellcheck disable=SC1091
|
|
echo \
|
|
"deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.asc] https://download.docker.com/linux/ubuntu \
|
|
$(. /etc/os-release && echo "${UBUNTU_CODENAME:-$VERSION_CODENAME}") stable" > /etc/apt/sources.list.d/docker.list
|
|
apt update -y
|
|
apt-cache madison docker-ce | awk '{ print $3 }'
|
|
verstr="$(apt-cache madison docker-ce | awk '{ print $3 }' | sort | grep -Ev ':(28|29)\.'| tail -n 1)"
|
|
../../contrib/apt-install -y "docker-ce=$verstr" "docker-ce-cli=$verstr"
|
|
|
|
# docker checkpoint and restore is an experimental feature
|
|
echo '{ "experimental": true }' > /etc/docker/daemon.json
|
|
service docker restart
|
|
|
|
CRIU_LOG='/criu.log'
|
|
mkdir -p /etc/criu
|
|
echo "log-file=$CRIU_LOG" > /etc/criu/runc.conf
|
|
|
|
# Test checkpoint/restore with action script
|
|
echo "action-script /usr/bin/true" | sudo tee /etc/criu/default.conf
|
|
|
|
export SKIP_CI_TEST=1
|
|
|
|
./run-ci-tests.sh
|
|
|
|
cd ../../
|
|
|
|
make install
|
|
|
|
docker info
|
|
|
|
criu --version
|
|
|
|
run_container () {
|
|
docker run \
|
|
--tmpfs /tmp \
|
|
--tmpfs /run \
|
|
--read-only \
|
|
--name cr \
|
|
--health-cmd='sleep 1' \
|
|
--health-interval=1s \
|
|
-d \
|
|
alpine \
|
|
/bin/sh -c 'i=0; while true; do echo $i; i=$(expr $i + 1); sleep 1; done'
|
|
}
|
|
|
|
wait_running () {
|
|
until [ "$(docker inspect -f '{{.State.Running}}' cr)" = "true" ]; do
|
|
sleep 1;
|
|
done;
|
|
}
|
|
|
|
wait_healthy () {
|
|
until [ "$(docker inspect -f '{{.State.Health.Status}}' cr)" = "healthy" ]; do
|
|
sleep 1;
|
|
done;
|
|
}
|
|
|
|
checkpoint_container () {
|
|
CHECKPOINT_NAME=$1
|
|
|
|
docker checkpoint create cr "$CHECKPOINT_NAME" &&
|
|
(docker exec cr true >> /dev/null 2>&1 && exit 1 || exit 0) &&
|
|
# wait for container to stop
|
|
docker wait cr
|
|
}
|
|
|
|
print_logs () {
|
|
cat "$(grep log 'log file:' | sed 's/log file:\s*//')" || true
|
|
docker logs cr || true
|
|
cat $CRIU_LOG || true
|
|
dmesg
|
|
docker ps
|
|
exit 1
|
|
}
|
|
|
|
declare -i max_restore_container_tries=3
|
|
|
|
restore_container () {
|
|
CHECKPOINT_NAME=$1
|
|
|
|
for i in $(seq $max_restore_container_tries); do
|
|
docker start --checkpoint "$CHECKPOINT_NAME" cr 2>&1 | tee log && break
|
|
|
|
# FIXME: There is a race condition in docker/containerd that causes
|
|
# docker to occasionally fail when starting a container from a
|
|
# checkpoint immediately after the checkpoint has been created.
|
|
# https://github.com/moby/moby/issues/42900
|
|
if grep -Eq '^Error response from daemon: failed to upload checkpoint to containerd: commit failed: content sha256:.*: already exists$' log; then
|
|
echo "Retry container restore: $i/$max_restore_container_tries"
|
|
sleep 1;
|
|
else
|
|
print_logs
|
|
fi
|
|
|
|
done
|
|
}
|
|
|
|
# Scenario: Create multiple containers and checkpoint and restore them once
|
|
for i in $(seq 10); do
|
|
run_container
|
|
wait_running
|
|
|
|
docker ps
|
|
checkpoint_container checkpoint
|
|
|
|
docker ps
|
|
restore_container checkpoint
|
|
|
|
docker ps
|
|
docker rm -f cr
|
|
done
|
|
|
|
# Scenario: Create container and checkpoint and restore it multiple times
|
|
run_container
|
|
wait_running
|
|
|
|
for i in $(seq 5); do
|
|
docker ps
|
|
checkpoint_container checkpoint"${i}"
|
|
|
|
docker ps
|
|
restore_container checkpoint"${i}"
|
|
|
|
# Wait for healthy state before creating another checkpoint
|
|
wait_healthy
|
|
done
|