项目背景
在内网NAS环境中，我部署了一台虚拟机，上面跑了十几个容器，随着应用变多，运维工作量也变得多起来，传统运维收到告警，还得开电脑，连线主机，看日志…
root@ubuntu01:~# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
89a81ba96d48 misskey/misskey:latest "/usr/bin/tini -- pn…" 29 hours ago Up 22 hours (healthy) 0.0.0.0:3500->3000/tcp, [::]:3500->3000/tcp misskey-web-1
f126c921c7aa postgres:15-alpine "docker-entrypoint.s…" 29 hours ago Up 22 hours (healthy) 5432/tcp misskey-db-1
ae9f056de32b redis:7-alpine "docker-entrypoint.s…" 29 hours ago Up 22 hours (healthy) 6379/tcp misskey-redis-1
88c496187043 soulter/astrbot:latest "python main.py" 2 days ago Up 22 hours 0.0.0.0:4000->6185/tcp, [::]:4000->6185/tcp astrbot
5cf8593834f9 yidadaa/chatgpt-next-web:latest "docker-entrypoint.s…" 3 days ago Up 22 hours 0.0.0.0:3900->3000/tcp, [::]:3900->3000/tcp chatgpt-next-web
1a68ed172722 xxnuo/mtranserver:latest "docker-entrypoint.s…" 4 days ago Up 22 hours 0.0.0.0:3700->8989/tcp, [::]:3700->8989/tcp mtranserver
d74ef3010948 justsong/one-api "/one-api" 5 days ago Up 22 hours 0.0.0.0:3800->3000/tcp, [::]:3800->3000/tcp oneapi
6db03f412728 privoce/vocechat-server:latest "/docker-entrypoint.…" 9 days ago Up 22 hours 0.0.0.0:3000->3000/tcp, [::]:3000->3000/tcp vocechat-server
a7a07b59fd30 yunyoujun/cook:latest "/docker-entrypoint.…" 2 weeks ago Up 22 hours 0.0.0.0:3600->80/tcp, [::]:3600->80/tcp cook
1f86c33ee759 henrygd/beszel-agent:latest "/agent" 3 weeks ago Up 22 hours beszel-agent
8f8419903140 excalidraw/excalidraw:latest "/docker-entrypoint.…" 3 weeks ago Up 22 hours (healthy) 0.0.0.0:3300->80/tcp, [::]:3300->80/tcp excalidraw
61b780d59789 superneed/global-radio:latest "/docker-entrypoint.…" 3 weeks ago Up 22 hours 0.0.0.0:3200->80/tcp, [::]:3200->80/tcp global-radio
73448c4dd1c1 systemsome/panhub:latest "docker-entrypoint.s…" 4 weeks ago Up 22 hours 0.0.0.0:3100->3000/tcp, [::]:3100->3000/tcp panhub
root@ubuntu01:~#
我们前面已经部署好的oneapi对接Glm及Agnes大模型，那能否通过AI来减少工作量？刚好宝塔面板新版本已经把AI嵌入到面板，且集成了非常多的技能及助手，直接让AI对接大模型，帮我们自动化运维，手机上和AI对话就可以帮我们把事情做完！最重要的是免费！！！！
正式部署
我们直接在现有业务的虚拟机上安装宝塔面板，安装过程会提示使用纯净系统安装，不支持已部署项目环境安装，放心直接强制安装即可，数据不会丢失
root@ubuntu01:/# mkdir /data/www
root@ubuntu01:/# ln -s /data/www www
root@ubuntu01:/# ls -al
total 2097236
drwxr-xr-x 22 root root 4096 Aug 1 14:37 .
drwxr-xr-x 22 root root 4096 Aug 1 14:37 ..
lrwxrwxrwx 1 root root 7 Apr 20 16:46 bin -> usr/bin
drwxr-xr-x 4 root root 4096 Jul 2 06:58 boot
dr-xr-xr-x 2 root root 4096 Apr 21 02:23 cdrom
drwxr-xr-x 17 root root 4096 Aug 1 14:37 data
drwxr-xr-x 21 root root 4160 Jul 31 16:59 dev
drwxr-xr-x 114 root root 4096 Jul 30 08:37 etc
drwxr-xr-x 3 root root 4096 Jul 1 09:24 home
lrwxrwxrwx 1 root root 7 Apr 20 16:46 lib -> usr/lib
lrwxrwxrwx 1 root root 9 Apr 20 16:46 lib64 -> usr/lib64
drwx------ 2 root root 16384 Jul 1 09:21 lost+found
drwxr-xr-x 2 root root 4096 Apr 21 02:05 media
drwxr-xr-x 2 root root 4096 Apr 21 02:05 mnt
drwxr-xr-x 3 root root 4096 Jul 1 10:45 opt
dr-xr-xr-x 265 root root 0 Jul 31 16:59 proc
drwx------ 7 root root 4096 Jul 31 17:08 root
drwxr-xr-x 38 root root 1080 Aug 1 14:33 run
lrwxrwxrwx 1 root root 8 Apr 20 16:46 sbin -> usr/sbin
drwxr-xr-x 2 root root 4096 Jul 1 09:30 snap
drwxr-xr-x 3 root root 4096 Jul 1 09:44 srv
-rw------- 1 root root 2147483648 Jul 1 09:24 swap.img
dr-xr-xr-x 13 root root 0 Aug 1 14:36 sys
drwxrwxrwt 13 root root 260 Aug 1 14:37 tmp
drwxr-xr-x 3 root root 4096 Jul 1 09:21 userap
drwxr-xr-x 12 root root 4096 Apr 21 02:05 usr
drwxr-xr-x 13 root root 4096 Jul 1 09:30 var
lrwxrwxrwx 1 root root 9 Aug 1 14:37 www -> /data/www
root@ubuntu01:/# cd /data/
root@ubuntu01:/data# wget -O install_panel.sh https://download.bt.cn/install/install_panel.sh && sudo bash install_panel.sh ed8484bec
--2026-08-01 14:39:17-- https://download.bt.cn/install/install_panel.sh
Resolving download.bt.cn (download.bt.cn)... 211.99.112.100, 2409:8c20:28c1:4::2
Connecting to download.bt.cn (download.bt.cn)|211.99.112.100|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 80645 (79K) [application/octet-stream]
Saving to: ‘install_panel.sh’
install_panel.sh 100%[=============================================================>] 78.75K --.-KB/s in 0.02s
2026-08-01 14:39:18 (4.82 MB/s) - ‘install_panel.sh’ saved [80645/80645]
cat: /etc/redhat-release: No such file or directory
+----------------------------------------------------------------------
| Bt-WebPanel FOR CentOS/Ubuntu/Debian
+----------------------------------------------------------------------
| Copyright © 2015-2099 BT-SOFT(http://www.bt.cn) All rights reserved.
+----------------------------------------------------------------------
| The WebPanel URL will be http://SERVER_IP:21863 when installed.
+----------------------------------------------------------------------
| 为了您的正常使用，请确保使用全新或纯净的系统安装宝塔面板，不支持已部署项目/环境的系统安装
+----------------------------------------------------------------------
Do you want to install Bt-Panel to the /www directory now?(y/n): y
----------------------------------------------------
检查已有其他Web/mysql环境，安装宝塔可能影响现有站点及数据
Web/mysql service is alreday installed,Can't install panel
----------------------------------------------------
已知风险/Enter yes to force installation
输入yes强制安装: yes
安装完成直接登录宝塔面板->Docker，可以看到已经自动识别到我们部署好的应用
我们直接点击AI->设置->添加自定义模型->设定模型为agnes-2.5-flash前提是已经在oneapi上分配好令牌，添加完成点击测试，通过记录
模型建议只保留agnes-2.5-flash,其他的都取消勾选，其他模型会消耗点数需要充值，agnes-2.5-flash已经够用了
回到AI对话窗口，可以看到第一次官方送了50个点数，这个先留着，后面遇到agnes-2.5-flash也搞不定的时候，再开启其他大模型再来用
直接来测试吧，让AI做一次深度系统巡检，对话窗口直接对AI聊天，反应特别快，报告也特别详细，点数也都没有消耗
提示CPU负载高，我们直接让他查询优化，和以前只会对答的AI不一样，他是可以直接帮你把事情做掉，动动手指就可以把事情做完了
写在后面
选择宝塔的原因，最重要的是他手机端APP是免费的，功能齐全，AI对话记录也是同步的，手机端和AI对话一下就可以把事情做掉了，当然，AI只是辅助，有时候也会抽风，大家可以引导他帮我们做正确的事情，还是不可以全部交给AI，大家可以自己部署探索体验哈
评论区