fast-reboot

#!/bin/bash

REBOOT_USER=$(logname)
REBOOT_TIME=$(date)
REBOOT_CAUSE_FILE="/var/cache/sonic/reboot-cause.txt"

# Check root privileges
if [[ "$EUID" -ne 0 ]]
then
  echo "This command must be run as root" >&2
  exit
fi


# Unload the previously loaded kernel if any loaded
if [[ "$(cat /sys/kernel/kexec_loaded)" -eq 1 ]]
then
  /sbin/kexec -u
fi

# Kernel and initrd image
NEXT_SONIC_IMAGE=$(sonic_installer list | grep "Next: " | cut -d ' ' -f 2)
if grep -q aboot_platform= /host/machine.conf; then
    IMAGE_PATH="/host/image-${NEXT_SONIC_IMAGE#SONiC-OS-}"
    KERNEL_IMAGE="$(ls $IMAGE_PATH/boot/vmlinuz-*)"
    BOOT_OPTIONS="$(cat "$IMAGE_PATH/kernel-cmdline" | tr '\n' ' ') fast-reboot"
elif grep -q onie_platform= /host/machine.conf; then
    KERNEL_OPTIONS=$(cat /host/grub/grub.cfg | sed "/$NEXT_SONIC_IMAGE'/,/}/"'!'"g" | grep linux)
    KERNEL_IMAGE="/host$(echo $KERNEL_OPTIONS | cut -d ' ' -f 2)"
    BOOT_OPTIONS="$(echo $KERNEL_OPTIONS | sed -e 's/\s*linux\s*/BOOT_IMAGE=/') fast-reboot"
else
    echo "Unknown bootloader. fast-reboot is not supported."
    exit 1
fi
INITRD=$(echo $KERNEL_IMAGE | sed 's/vmlinuz/initrd.img/g')

sonic_asic_type=$(sonic-cfggen -y /etc/sonic/sonic_version.yml -v asic_type)

# Install new FW for mellanox platforms before control plane goes down
# So on boot switch will not spend time to upgrade FW increasing the CP downtime
if [[ "$sonic_asic_type" == "mellanox" ]];
then
    CURRENT_SONIC_IMAGE=$(sonic_installer list | grep "Current: " | cut -d ' ' -f 2)
    if [[ "${CURRENT_SONIC_IMAGE}" != "${NEXT_SONIC_IMAGE}" ]]; then
        echo "Prepare ASIC to fast reboot: install new FW if requiered"
        NEXT_IMAGE_FS_PATH="/host/image-${NEXT_SONIC_IMAGE#SONiC-OS-}/fs.squashfs"
        FS_MOUNTPOINT="/tmp/image-${NEXT_SONIC_IMAGE#SONiC-OS-}-fs"
        mkdir -p "$FS_MOUNTPOINT"
        mount -t squashfs "$NEXT_IMAGE_FS_PATH" "$FS_MOUNTPOINT"
        /usr/bin/mlnx-fw-upgrade.sh "$FS_MOUNTPOINT/etc/mlnx/fw-SPC.mfa"
        EXIT_CODE=$?
        umount "$FS_MOUNTPOINT"
        if [[ $EXIT_CODE != 0 ]]; then
            echo "Failed to burn FW"
            exit 1
        fi
    fi
fi

# Load kernel into the memory
/sbin/kexec -l "$KERNEL_IMAGE" --initrd="$INITRD" --append="$BOOT_OPTIONS"

# Dump the ARP and FDB tables to files also as default routes for both IPv4 and IPv6
# into /host/fast-reboot
mkdir -p /host/fast-reboot
/usr/bin/fast-reboot-dump.py -t /host/fast-reboot

# Kill bgpd to start the bgp graceful restart procedure
docker exec -i bgp killall -9 zebra
docker exec -i bgp killall -9 bgpd

# Kill lldp, otherwise it sends informotion about reboot
docker kill lldp > /dev/null

# Kill teamd, otherwise it gets down all LAGs
docker kill teamd > /dev/null

# syncd graceful stop is supported only for Broadcom platforms only for now
if [[ "$sonic_asic_type" = 'broadcom' ]];
then
    # Gracefully stop syncd
    docker exec -i syncd /usr/bin/syncd_request_shutdown --cold > /dev/null

    # Check that syncd was stopped
    while docker top syncd | grep -q /usr/bin/syncd
    do
          sleep 0.1
    done
fi

# Kill other containers to make the reboot faster
docker ps -q | xargs docker kill > /dev/null

# Stop the docker container engine. Otherwise we will have a broken docker storage
systemctl stop docker.service

# Stop opennsl modules for Broadcom platform
if [[ "$sonic_asic_type" = 'broadcom' ]];
then
  service_name=$(systemctl list-units --plain --no-pager --no-legend --type=service | grep opennsl | cut -f 1 -d' ')
  systemctl stop "$service_name"
fi

# Stop kernel modules for Nephos platform
if [[ "$sonic_asic_type" = 'nephos' ]];
then
  systemctl stop nps-modules-`uname -r`.service
fi

# Update the reboot cause file to reflect that user issued 'fast-reboot' command
# Upon next boot, the contents of this file will be used to determine the
# cause of the previous reboot
echo "User issued 'fast-reboot' command [User: ${REBOOT_USER}, Time: ${REBOOT_TIME}]" > ${REBOOT_CAUSE_FILE}

# Wait until all buffers synced with disk
sync
sleep 1
sync

# Reboot: explicity call Linux native reboot under sbin
echo "Rebooting to $NEXT_SONIC_IMAGE..."
exec /sbin/reboot

# Should never reach here
echo "fast-reboot failed!" >&2
exit 1