Move teamd warm reboot code to service script (#5163)

Summary: Move teamd functions to a new service script

Motivation: To segregate teamd functions in one common place. fast-reboot script calls teamd functions that should ideally be replaced by a simple call to a service script.
 
Changes: New teamd service script and path modification from /usr/bin/teamd.sh to /usr/local/bin/teamd.sh
fast-reboot script (in sonic-utilities) modification (to use new teamd.sh to stop teamd) should follow soon after this change.

Verification: VS image tests.

Signed-off-by: Vaibhav Hemant Dixit <vaibhav.dixit@microsoft.com>

Co-authored-by: heidi.ou@alibaba-inc.com <heidi.ou@alibaba-inc.com>
Co-authored-by: Ying Xie <ying.xie@microsoft.com>
This commit is contained in:
heidinet2007 2020-11-13 13:34:18 -08:00 committed by GitHub
parent 56fa3cf494
commit 7c17c58b83
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 113 additions and 4 deletions

View File

@ -9,9 +9,9 @@ StartLimitBurst=3
[Service]
User={{ sonicadmin_user }}
ExecStartPre=/usr/bin/{{docker_container_name}}.sh start{% if multi_instance == 'true' %} %i{% endif %}
ExecStart=/usr/bin/{{docker_container_name}}.sh wait{% if multi_instance == 'true' %} %i{% endif %}
ExecStop=/usr/bin/{{docker_container_name}}.sh stop{% if multi_instance == 'true' %} %i{% endif %}
ExecStartPre=/usr/local/bin/{{docker_container_name}}.sh start{% if multi_instance == 'true' %} %i{% endif %}
ExecStart=/usr/local/bin/{{docker_container_name}}.sh wait{% if multi_instance == 'true' %} %i{% endif %}
ExecStop=/usr/local/bin/{{docker_container_name}}.sh stop{% if multi_instance == 'true' %} %i{% endif %}
Restart=always
RestartSec=30

View File

@ -565,12 +565,13 @@ sudo LANG=C chroot $FILESYSTEM_ROOT fuser -km /sys || true
sudo LANG=C chroot $FILESYSTEM_ROOT umount -lf /sys
{% endif %}
# Copy service scripts (swss, syncd, bgp, radv)
# Copy service scripts (swss, syncd, bgp, teamd, radv)
sudo LANG=C cp $SCRIPTS_DIR/swss.sh $FILESYSTEM_ROOT/usr/local/bin/swss.sh
sudo LANG=C cp $SCRIPTS_DIR/syncd.sh $FILESYSTEM_ROOT/usr/local/bin/syncd.sh
sudo LANG=C cp $SCRIPTS_DIR/syncd_common.sh $FILESYSTEM_ROOT/usr/local/bin/syncd_common.sh
sudo LANG=C cp $SCRIPTS_DIR/gbsyncd.sh $FILESYSTEM_ROOT/usr/local/bin/gbsyncd.sh
sudo LANG=C cp $SCRIPTS_DIR/bgp.sh $FILESYSTEM_ROOT/usr/local/bin/bgp.sh
sudo LANG=C cp $SCRIPTS_DIR/teamd.sh $FILESYSTEM_ROOT/usr/local/bin/teamd.sh
sudo LANG=C cp $SCRIPTS_DIR/radv.sh $FILESYSTEM_ROOT/usr/local/bin/radv.sh
# Copy sonic-netns-exec script

108
files/scripts/teamd.sh Executable file
View File

@ -0,0 +1,108 @@
#!/bin/bash
function debug()
{
/usr/bin/logger $1
/bin/echo `date` "- $1" >> ${DEBUGLOG}
}
function check_warm_boot()
{
SYSTEM_WARM_START=`$SONIC_DB_CLI STATE_DB hget "WARM_RESTART_ENABLE_TABLE|system" enable`
SERVICE_WARM_START=`$SONIC_DB_CLI STATE_DB hget "WARM_RESTART_ENABLE_TABLE|${SERVICE}" enable`
if [[ x"$SYSTEM_WARM_START" == x"true" ]] || [[ x"$SERVICE_WARM_START" == x"true" ]]; then
WARM_BOOT="true"
else
WARM_BOOT="false"
fi
}
function validate_restore_count()
{
if [[ x"$WARM_BOOT" == x"true" ]]; then
RESTORE_COUNT=`$SONIC_DB_CLI STATE_DB hget "WARM_RESTART_TABLE|${SERVICE}" restore_count`
# We have to make sure db data has not been flushed.
if [[ -z "$RESTORE_COUNT" ]]; then
WARM_BOOT="false"
fi
fi
}
function check_fast_boot ()
{
if [[ $($SONIC_DB_CLI STATE_DB GET "FAST_REBOOT|system") == "1" ]]; then
FAST_BOOT="true"
else
FAST_BOOT="false"
fi
}
start() {
debug "Starting ${SERVICE}$DEV service..."
check_warm_boot
validate_restore_count
check_fast_boot
debug "Warm boot flag: ${SERVICE}$DEV ${WARM_BOOT}."
debug "Fast boot flag: ${SERVICE}$DEV ${Fast_BOOT}."
# start service docker
/usr/bin/${SERVICE}.sh start $DEV
debug "Started ${SERVICE}$DEV service..."
}
wait() {
/usr/bin/${SERVICE}.sh wait $DEV
}
stop() {
debug "Stopping ${SERVICE}$DEV service..."
check_warm_boot
check_fast_boot
debug "Warm boot flag: ${SERVICE}$DEV ${WARM_BOOT}."
debug "Fast boot flag: ${SERVICE}$DEV ${FAST_BOOT}."
if [[ x"$WARM_BOOT" == x"true" ]]; then
# Send USR1 signal to all teamd instances to stop them
# It will prepare teamd for warm-reboot
# Note: We must send USR1 signal before syncd, because it will send the last packet through CPU port
docker exec -i ${SERVICE}$DEV pkill -USR1 ${SERVICE} > /dev/null || [ $? == 1 ]
elif [[ x"$FAST_BOOT" == x"true" ]]; then
# Kill teamd processes inside of teamd container with SIGUSR2 to allow them to send last LACP frames
# We call `docker kill teamd` to ensure the container stops as quickly as possible,
# Note: teamd must be killed before syncd, because it will send the last packet through CPU port
docker exec -i ${SERVICE}$DEV pkill -USR2 ${SERVICE} || [ $? == 1 ]
while docker exec -i ${SERVICE}$DEV pgrep ${SERVICE} > /dev/null; do
sleep 0.05
done
docker kill ${SERVICE}$DEV &> /dev/null || debug "Docker ${SERVICE}$DEV is not running ($?) ..."
fi
/usr/bin/${SERVICE}.sh stop $DEV
debug "Stopped ${SERVICE}$DEV service..."
}
DEV=$2
SERVICE="teamd"
DEBUGLOG="/tmp/teamd-debug$DEV.log"
NAMESPACE_PREFIX="asic"
if [ "$DEV" ]; then
NET_NS="$NAMESPACE_PREFIX$DEV" #name of the network namespace
SONIC_DB_CLI="sonic-db-cli -n $NET_NS"
else
SONIC_DB_CLI="sonic-db-cli"
fi
case "$1" in
start|wait|stop)
$1
;;
*)
echo "Usage: $0 {start|wait|stop}"
exit 1
;;
esac