feature: P-Agent 安装 IPv6 场景适配 (closed TencentBlueKing#1290)

apps/backend/agent/solution_maker.py

@@ -710,10 +710,7 @@ def get_run_cmd_base_params(self) -> typing.List[str]:
             # 代理机器配置
             f"-HPP '{settings.BK_NODEMAN_NGINX_PROXY_PASS_PORT}'",
             # 代理机器主机信息
-            f"-I {self.gse_servers_info['jump_server'].inner_ip}",
-            f"-I6 {self.gse_servers_info['jump_server'].inner_ipv6}"
-            if self.gse_servers_info["jump_server"].inner_ipv6
-            else "",
+            f"-I {self.gse_servers_info['jump_server'].inner_ip or self.gse_servers_info['jump_server'].inner_ipv6}",
         ]
 
         # 通道特殊配置

apps/backend/components/collections/agent_new/install.py

@@ -395,7 +395,7 @@ def execute_job_commands(self, sub_inst_id, installation_tool: InstallationTools
         )
         self.log_info(
             sub_inst_ids=sub_inst_id,
-            log_content=_("已选择 {inner_ip} 作为本次安装的跳板机").format(inner_ip=jump_server.inner_ip),
+            log_content=_("已选择 {inner_ip} 作为本次安装的跳板机").format(inner_ip=jump_server.inner_ip or jump_server.inner_ipv6),
         )
         path = os.path.join(settings.BK_SCRIPTS_PATH, constants.SetupScriptFileName.SETUP_PAGENT_PY.value)
         with open(path, encoding="utf-8") as fh:
@@ -453,8 +453,8 @@ def execute_job_commands(self, sub_inst_id, installation_tool: InstallationTools
                         "2. Proxy是否已正确完成所有安装步骤且状态正常。 \n"
                         "3. 点击上面链接跳转到作业平台查看任务执行情况。\n"
                     ).format(
-                        host_inner_ip=host.inner_ip,
-                        jump_server_ip=jump_server.inner_ip,
+                        host_inner_ip=host.inner_ip or host.inner_ipv6,
+                        jump_server_ip=jump_server.inner_ip or host.inner_ipv6,
                         download_port=settings.BK_NODEMAN_NGINX_DOWNLOAD_PORT,
                         proxy_pass_port=settings.BK_NODEMAN_NGINX_PROXY_PASS_PORT,
                     ),

apps/node_man/serializers/job.py

@@ -263,7 +263,7 @@ def validate(self, attrs):
         attrs["bk_host_ids"] = bk_host_ids
         attrs["bk_biz_scope"] = bk_biz_scope
 
-        if attrs["job_type"] not in ["REINSTALL_PROXY", "UPGRADE_PROXY"]:
+        if attrs["job_type"] not in ["REINSTALL_PROXY", "UPGRADE_PROXY", "UNINSTALL_PROXY"]:
             return attrs
 
         # 如果开启业务灰度，安装新版本 Agent

dev_log/2.2.32/crayon_202211292126.yaml

@@ -0,0 +1,3 @@
+---
+feature:
+  - "P-Agent 安装 IPv6 场景适配 (closed #1290)"

script_tools/agent_tools/agent2/setup_agent.bat

@@ -8,6 +8,7 @@ rem DEFAULT DEFINITION
 set PKG_NAME=
 set CLOUD_ID=
 set UNINSTALL=1
+set UNREGISTER_AGENT_ID=1
 set BACKUP_CONFIG_FILE=procinfo.json
 
 :CheckOpts
@@ -41,6 +42,7 @@ if "%1" EQU "-K" (set TRACKER_PORT=%~2) && shift && shift && goto CheckOpts
 if "%1" EQU "-U" (set INSTALL_USER=%~2) && shift && shift && goto CheckOpts
 if "%1" EQU "-P" (set INSTALL_PASSWORD=%~2) && shift && shift && goto CheckOpts
 if "%1" EQU "-R" (set UNINSTALL=0) && shift && shift && goto CheckOpts
+if "%1" EQU "-F" (set UNREGISTER_AGENT_ID=0) && shift && goto CheckOpts
 if "%1" NEQ "" echo Invalid option: "%1" && goto :EOF && exit /B 1
 rem if "%1" EQU "-R" goto remove_agent_pro
 
@@ -187,7 +189,7 @@ goto :EOF
     ping -n 1 127.0.0.1 >nul 2>&1
     for %%a in (%CLUSTER_PORT%,%DATA_PORT%) do (
         ping -n 1 127.0.0.1 >nul 2>&1
-        netstat -an | findstr /r /C:":%%a *ESTABLISHED" 
+        netstat -an | findstr /r /C:":%%a *ESTABLISHED"
         if !errorlevel! equ 0 (
             for /F "tokens=*" %%i in ('netstat -an ^| findstr /r /C:":%%a *ESTABLISHED"') do (set conninfo=%%i)
             call :print_check_deploy_result INFO check_deploy_result - "%%a connect to gse server success" 1>nul 2>&1
@@ -219,7 +221,7 @@ goto :EOF
                     call :multi_report_step_status
                 ) else (
                     call :print WARN check_env - "connect to upstream cluster server: %%p %CLUSTER_PORT% failed"
-                    set network_not_reachable=%network_not_reachable% %%p to %CLUSTER_PORT% 
+                    set network_not_reachable=%network_not_reachable% %%p to %CLUSTER_PORT%
                     call :multi_report_step_status
                 )
             )
@@ -228,7 +230,7 @@ goto :EOF
     if "!network_not_reachable!" == "" (
         echo "reachable"
     ) else (
-        call :print FAIL check_env FAILED "%network_not_reachable% is not reachable"    
+        call :print FAIL check_env FAILED "%network_not_reachable% is not reachable"
         call :multi_report_step_status
         exit /b 1
     )
@@ -247,7 +249,7 @@ goto :EOF
                     call :multi_report_step_status
                 ) else (
                     call :print WARN check_env - "connect to upstream data server: %%p %DATA_PORT% failed"
-                    set data_network_not_reachable=%data_network_not_reachable% %%p to %DATA_PORT% 
+                    set data_network_not_reachable=%data_network_not_reachable% %%p to %DATA_PORT%
                     call :multi_report_step_status
                 )
             )
@@ -256,7 +258,7 @@ goto :EOF
     if "!data_network_not_reachable!" == "" (
         echo "reachable"
     ) else (
-        call :print FAIL check_env FAILED "%data_network_not_reachable% is not reachable"    
+        call :print FAIL check_env FAILED "%data_network_not_reachable% is not reachable"
         call :multi_report_step_status
         exit /b 1
     )
@@ -276,7 +278,7 @@ goto :EOF
                         call :multi_report_step_status
                     ) else (
                         call :print WARN check_env - "connect to upstream file server: %%p %%a failed"
-                        set file_network_not_reachable=%file_network_not_reachable% %%p to %%a 
+                        set file_network_not_reachable=%file_network_not_reachable% %%p to %%a
                         call :multi_report_step_status
                     )
                 )
@@ -286,7 +288,7 @@ goto :EOF
     if "!file_network_not_reachable!" == "" (
         echo "reachable"
     ) else (
-        call :print FAIL check_env FAILED "%file_network_not_reachable% is not reachable"    
+        call :print FAIL check_env FAILED "%file_network_not_reachable% is not reachable"
         call :multi_report_step_status
         exit /b 1
     )
@@ -312,7 +314,7 @@ goto :EOF
                     call :multi_report_step_status
                 ) else (
                     call :print WARN check_env - "gse server %%p %CLUSTER_PORT% can not reachable"
-                    set network_not_reachable=%network_not_reachable% %%p to %CLUSTER_PORT% 
+                    set network_not_reachable=%network_not_reachable% %%p to %CLUSTER_PORT%
                     call :multi_report_step_status
                 )
             )
@@ -321,7 +323,7 @@ goto :EOF
     if "!network_not_reachable!" == "" (
         echo "reachable"
     ) else (
-        call :print FAIL check_env FAILED "%network_not_reachable% is not reachable"    
+        call :print FAIL check_env FAILED "%network_not_reachable% is not reachable"
         call :multi_report_step_status
         exit /b 1
     )
@@ -339,7 +341,7 @@ goto :EOF
                     call :multi_report_step_status
                 ) else (
                     call :print WARN check_env - "gse server %%p %DATA_PORT% can not reachable"
-                    set network_not_reachable=%network_not_reachable% %%p to %DATA_PORT% 
+                    set network_not_reachable=%network_not_reachable% %%p to %DATA_PORT%
                     call :multi_report_step_status
                 )
             )
@@ -348,7 +350,7 @@ goto :EOF
     if "!network_not_reachable!" == "" (
         echo "reachable"
     ) else (
-        call :print FAIL check_env FAILED "%network_not_reachable% is not reachable"    
+        call :print FAIL check_env FAILED "%network_not_reachable% is not reachable"
         call :multi_report_step_status
         exit /b 1
     )
@@ -366,7 +368,7 @@ goto :EOF
                         call :print INFO check_env - "gse server %%p %%a is reachable"
                         call :multi_report_step_status
                     ) else (
-                        set network_not_reachable=%network_not_reachable% %%p to %%a 
+                        set network_not_reachable=%network_not_reachable% %%p to %%a
                         call :print INFO check_env - "gse server %%p %%a can not reachable"
                         call :multi_report_step_status
                     )
@@ -1071,7 +1073,7 @@ goto :EOF
 goto :EOF
 
 :check_uninstall_result
-    
+
     set service=1
     sc query gse_agent_daemon_%service_id% | findstr /C:"does not exist" 1>nul 2>&1
     if %errorlevel% equ 0 (
@@ -1103,6 +1105,11 @@ goto :EOF
         call :multi_report_step_status
         exit /b 1
     )
+
+    if %UNREGISTER_AGENT_ID% EQU 0 (
+        call :unregister_agent_id SKIP
+    )
+
     %gse_winagent_home%\\agent\\bin\\gse_agent.exe --agent-id -f %GSE_AGENT_ETC_DIR%\\gse_agent.conf
     if %errorlevel% EQU 0 (
         call :re_register_agent_id
@@ -1214,6 +1221,7 @@ goto :EOF
 goto :EOF
 
 :unregister_agent_id
+    (set skip=%1)
     if NOT EXIST %gse_winagent_home%\\agent\\bin\\gse_agent.exe (
         call :print FAIL unregister_agent_id FAILED "%gse_winagent_home%\\agent\\bin\\gse_agent.exe not exists"
         call :multi_report_step_status
@@ -1232,12 +1240,17 @@ goto :EOF
         %gse_winagent_home%\\agent\\bin\\gse_agent.exe --unregister
     )
     if %errorlevel% EQU 0 (
-        call :print INFO register_agent_id - "unregister agent id success"
+        call :print INFO unregister_agent_id - "unregister agent id success"
         call :multi_report_step_status
     ) else (
-        call :print FAIL register_agent_id FAILED "unregister agent id failed"
-        call :multi_report_step_status
-        exit /b 1
+        if "%skip%" == "SKIP": (
+            call :print WARN unregister_agent_id - "unregister agent id failed, but skip it"
+            call :multi_report_step_status
+        ) else (
+            call :print FAIL unregister_agent_id FAILED "unregister agent id failed"
+            call :multi_report_step_status
+            exit /b 1
+        )
     )
 goto :EOF
 
@@ -1350,6 +1363,7 @@ goto :EOF
     echo -O CLUSTER_PORT
     echo -E FILE_SVR_PORT
     echo -R UNINSTALL_AGENT
+    echo -F UNREGISTER_AGENT_ID [optional]
 goto :EOF
 
 :EOF

script_tools/agent_tools/agent2/setup_agent.sh

@@ -331,6 +331,12 @@ register_agent_id () {
         fail register_agent_id FAILED "gse_agent file not exists in $AGENT_SETUP_PATH/bin"
     fi
 
+
+    if [[ "${UNREGISTER_AGENT_ID}" == "TRUE" ]]; then
+        log register_agent_id - "trying to unregister agent id"
+        unregister_agent_id SKIP
+    fi
+
     log register_agent_id  - "trying to register agent id"
     if [ -f "${GSE_AGENT_CONFIG_PATH}" ]; then
         registe_result=$($AGENT_SETUP_PATH/bin/gse_agent -f "${GSE_AGENT_CONFIG_PATH}" --register 2>&1)
@@ -346,6 +352,7 @@ register_agent_id () {
 }
 
 unregister_agent_id () {
+    local skip="$1"
     log unregister_agent_id - "trying to unregister agent id"
     if [ -f "$AGENT_SETUP_PATH/bin/gse_agent" ]; then
         if [ -f "${GSE_AGENT_CONFIG_PATH}" ]; then
@@ -357,7 +364,11 @@ unregister_agent_id () {
         if [[ $? -eq 0 ]]; then
             log unregister_agent_id SUCCESS "unregister agent id succeed"
         else
-            fail unregister_agent_id FAILED "unregister agent id failed, error: ${unregister_agent_id_result}"
+            if [[ "${skip}" == "SKIP" ]]; then
+                warn unregister_agent_id - "unregister agent id failed, but skip it. error: ${unregister_agent_id_result}"
+            else
+                fail unregister_agent_id FAILED "unregister agent id failed, error: ${unregister_agent_id_result}"
+            fi
         fi
     else
         log unregister_agent_id FAILED "gse_agent file not exists in $AGENT_SETUP_PATH/bin"
@@ -756,6 +767,7 @@ _help () {
     echo "  -S BT_PORT_START"
     echo "  -Z BT_PORT_END"
     echo "  -K TRACKER_PORT"
+    echo "  -F UNREGISTER_AGENT_ID [optional]"
 
     exit 0
 }
@@ -783,6 +795,7 @@ AGENT_SETUP_PATH="/usr/local/gse/${NODE_TYPE}"
 CURR_PID=$$
 OVERIDE=false
 REMOVE=false
+UNREGISTER_AGENT_ID=false
 CALLBACK_URL=
 AGENT_PID=
 DEBUG=
@@ -792,7 +805,7 @@ LOG_RPT_CNT=0
 BULK_LOG_SIZE=3
 
 # main program
-while getopts n:t:I:i:l:s:uc:r:x:p:e:a:k:N:v:oT:RDO:E:A:V:B:S:Z:K: arg; do
+while getopts n:t:I:i:l:s:uc:r:x:p:e:a:k:N:v:oT:RDO:E:A:V:B:S:Z:K:F arg; do
     case $arg in
         n) NAME="$OPTARG" ;;
         t) VERSION="$OPTARG" ;;
@@ -821,6 +834,7 @@ while getopts n:t:I:i:l:s:uc:r:x:p:e:a:k:N:v:oT:RDO:E:A:V:B:S:Z:K: arg; do
         S) BT_PORT_START=$OPTARG ;;
         Z) BT_PORT_END=$OPTARG ;;
         K) TRACKER_PORT=$OPTARG ;;
+        F) UNREGISTER_AGENT_ID=TRUE ;;
         *)  _help ;;
     esac
 done

script_tools/agent_tools/agent2/setup_proxy.sh

@@ -185,7 +185,7 @@ is_port_connected_by_pid () {
         sleep 1
         stat -L -c %i /proc/"$pid"/fd/* 2>/dev/null \
             | grep -qwFf - \
-                <( awk -v p="$port" 'BEGIN{ check=sprintf(":%04X01$", p)} $3$4 ~ check {print $10}' /proc/net/tcp) \
+                <( awk -v p="$port" 'BEGIN{ check=sprintf(":%04X01$", p)} $3$4 ~ check {print $10}' /proc/net/tcp*) \
                 && return 0
     done
     return 1

script_tools/setup_pagent.py

@@ -6,17 +6,19 @@
 
 import argparse
 import base64
+import ipaddress
 import json
 import os
 import re
 import socket
 import sys
 import time
 import traceback
+from functools import partial
 from io import StringIO
 from pathlib import Path
 from subprocess import Popen
-from typing import Any, Dict, List
+from typing import Any, Dict, List, Optional
 
 PRIVATE_KEY_MERGED_TEXT = """
 %%PRIVATE_KEY_MERGED_TEXT%%
@@ -45,6 +47,29 @@
 JOB_PRIVATE_KEY_RE = re.compile(r"^(-{5}BEGIN .*? PRIVATE KEY-{5})(.*?)(-{5}END .*? PRIVATE KEY-{5}.?)$")
 
 
+def is_ip(ip: str, _version: Optional[int] = None) -> bool:
+    """
+    判断是否为合法 IP
+    :param ip:
+    :param _version: 是否为合法版本，缺省表示 both
+    :return:
+    """
+    try:
+        ip_address = ipaddress.ip_address(ip)
+    except ValueError:
+        return False
+    if _version is None:
+        return True
+    return ip_address.version == _version
+
+
+# 判断是否为合法 IPv6
+is_v6 = partial(is_ip, _version=6)
+
+# 判断是否为合法 IPv4
+is_v4 = partial(is_ip, _version=4)
+
+
 def arg_parser() -> argparse.ArgumentParser:
     """Commandline argument parser"""
     parser = argparse.ArgumentParser(description="p-agent setup scripts")
@@ -277,7 +302,7 @@ def execute_shell_solution(
 
 
 def is_port_listen(ip: str, port: int) -> bool:
-    s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    s = socket.socket((socket.AF_INET, socket.AF_INET6)[is_v6(ip)], socket.SOCK_STREAM)
     r = s.connect_ex((ip, port))
 
     if r == 0: