mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-12-24 13:28:13 +08:00
[PD Disaggregation] support DP via v1 router and decouple DP and EP (#5197)
* [fix] support DP via v1 router and decouple DP and EP * [fix] fix scripts * [fix] reset model path * [fix] dp use get_output_ep, fix router port type, update scripts * [merge] merge with latest code * [chore] remove some debug log * [fix] fix code style check * [fix] fix test_multi_api_server for log_dir name * [chore] reduce logs * Apply suggestions from code review Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
This commit is contained in:
@@ -1,8 +1,16 @@
|
||||
#!/bin/bash
|
||||
|
||||
is_port_free() {
|
||||
local port=$1
|
||||
if ss -ltn | awk '{print $4}' | grep -q ":${port}$"; then
|
||||
return 1 # Port is occupied
|
||||
fi
|
||||
return 0 # Port is free
|
||||
}
|
||||
|
||||
check_ports() {
|
||||
for port in "$@"; do
|
||||
if ss -tuln | grep -q ":$port "; then
|
||||
if ! is_port_free $port; then
|
||||
echo "❌ Port $port is already in use"
|
||||
return 1
|
||||
fi
|
||||
@@ -11,14 +19,79 @@ check_ports() {
|
||||
}
|
||||
|
||||
wait_for_health() {
|
||||
local server_port=$1
|
||||
IFS=',' read -r -a server_ports <<< "$1"
|
||||
local num_ports=${#server_ports[@]}
|
||||
local total_lines=$((num_ports + 1))
|
||||
local first_run=true
|
||||
local GREEN='\033[0;32m'
|
||||
local RED='\033[0;31m'
|
||||
local NC='\033[0m' # No Color
|
||||
local start_time=$(date +%s)
|
||||
|
||||
while true; do
|
||||
status_code=$(curl -s -o /dev/null -w "%{http_code}" "http://0.0.0.0:${server_port}/health" || echo "000")
|
||||
if [ "$status_code" -eq 200 ]; then
|
||||
local all_ready=true
|
||||
for port in "${server_ports[@]}"; do
|
||||
status_code=$(curl -s --max-time 1 -o /dev/null -w "%{http_code}" "http://0.0.0.0:${port}/health" || echo "000")
|
||||
if [ "$status_code" -eq 200 ]; then
|
||||
printf "Port %s: ${GREEN}[OK] 200${NC}\033[K\n" "$port"
|
||||
else
|
||||
all_ready=false
|
||||
printf "Port %s: ${RED}[WAIT] %s${NC}\033[K\n" "$port" "$status_code"
|
||||
fi
|
||||
done
|
||||
cur_time=$(date +%s)
|
||||
if [ "$all_ready" = "true" ]; then
|
||||
echo "All services are ready! [$((cur_time-start_time))s]"
|
||||
break
|
||||
else
|
||||
echo "Service not ready. Retrying in 4s..."
|
||||
sleep 4
|
||||
fi
|
||||
else
|
||||
echo "Waiting for services... [$((cur_time-start_time))s]"
|
||||
printf "\033[%dA" "$total_lines" # roll back cursor
|
||||
sleep 1
|
||||
fi
|
||||
done
|
||||
}
|
||||
|
||||
get_free_ports() {
|
||||
free_ports_num=${1:-1}
|
||||
start_port=${2:-8000}
|
||||
end_port=${3:-9000}
|
||||
|
||||
free_ports=()
|
||||
if [[ ! -n ${free_ports_num} || "${free_ports_num}" -le 0 ]]; then
|
||||
log_warn "param can't be empty, and should > 0"
|
||||
echo ${free_ports[@]}
|
||||
return 1
|
||||
fi
|
||||
|
||||
used_ports1=$(netstat -an | grep -E "(0.0.0.0|127.0.0.1|${POD_IP}|tcp6)" | awk '{n=split($4,a,":"); if(a[n]~/^[0-9]+$/) print a[n];}' | sort -u)
|
||||
used_ports2=$(netstat -an | grep -E "(0.0.0.0|127.0.0.1|${POD_IP}|tcp6)" | awk '{n=split($5,a,":"); if(a[n]~/^[0-9]+$/) print a[n];}' | sort -u)
|
||||
all_used_ports=$(printf "%s\n" "${used_ports1}" "${used_ports2}" | sort -u)
|
||||
|
||||
# Generate random number between 0 and 32767
|
||||
random_num=$(( RANDOM ))
|
||||
port=$(( random_num % (end_port - start_port + 1) + start_port ))
|
||||
|
||||
while true; do
|
||||
(( port++ ))
|
||||
if [[ ${port} -ge ${end_port} ]]; then
|
||||
port=${start_port}
|
||||
fi
|
||||
|
||||
if [[ "${all_used_ports[@]}" =~ "${port}" ]]; then
|
||||
continue
|
||||
fi
|
||||
|
||||
if is_port_free ${port}; then
|
||||
free_ports+=("${port}")
|
||||
(( free_ports_num-- ))
|
||||
if [[ ${free_ports_num} = 0 ]]; then
|
||||
break
|
||||
fi
|
||||
fi
|
||||
|
||||
done
|
||||
|
||||
# echo ${free_ports[@]}
|
||||
IFS=',' && echo "${free_ports[*]}"
|
||||
return 0
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user