Files
OJ2/docker/compose.school.yml
T
xuyueandClaude Opus 5.5 37e971b6b4 refactor(判题机): 换成 go-judge,空闲内存从几百 MB 降到 16 MB
原来的 JudgeServer 按核数起 gunicorn worker、每份提交再按核数开进程池,空闲内存跟着核数走
(2 核 52MB、16 核 305MB)。换成 go-judge 后空闲 16MB,同一批提交判题中 80–100MB(原来约 370MB)。

go-judge 只管在沙箱里跑命令,编译、逐点运行、比对输出、定状态码搬到 judge/sandbox.ts,
口径逐条照搬 JudgeServer,出参形状不变。拿生产库真实提交新旧逐点对比:C 2000 份 1989 一致、
C++ 891 份 890 一致、Python 2000 份全一致;不一致的是 system("pause")(现在不再判运行时错误)、
未定义行为、endl 死循环打印(运行时错误变运行超时),见 docker/judge/README.md。

- 判题容器改成 privileged(go-judge 要自己建 cgroup),镜像 oj2-judge-3
- 心跳改由 worker 每 5 秒问判题机 /version 代记,删掉 /api/judge-server/heartbeat
- 前端统一用 stripJudgePath 去掉判题机工作目录(新的 /w/ 和旧的 /judger/run/)
- 提交前编译检查只编译不运行

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-26 15:03:57 +08:00

117 lines
4.2 KiB
YAML
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# 机房。**这里没有数据库** —— 连的是服务器那台的 5445(见 compose.debian.yml)。
#
# docker compose -f docker/compose.school.yml --env-file docker/.env.school up -d --build
#
# 两个站点共用一个库,但各有各的 Redis、判题沙箱和后端。这意味着:
#
# - 判题队列是每站独立的(BullMQ 在本地 Redis 上),学生在哪边提交就在哪边判,
# 和旧后端的 Dramatiq 拓扑一致;
# - WebSocket 推送也走本地 Redis pub/sub,所以只推得到连在**本站**的学生。
# 旧后端的 Channels 也是这样,不是回归;
# - **切换那天两边都要切。** 只切一边的话,另一边的旧后端仍在读写同一个库,
# 而库结构已经按新后端迁过了。
#
# ⚠️ `DATA_DIR` 默认值 `../data` 解析出来是 **`OJ2/data`**,不是机房部署目录的 `data/`。
# 测试点(`data/backend/test_case`)和题面图片(`data/backend/public/upload`)都在旧目录里,
# 不设 `DATA_DIR` 就会挂一堆空目录上去:判题全错、图片 404。**沿用旧数据必须设它。**
services:
oj-redis:
image: redis:8-alpine
container_name: oj-redis
restart: always
volumes:
- ${DATA_DIR:-../data}/redis:/data
healthcheck:
test: ["CMD", "redis-cli", "ping"]
interval: 5s
timeout: 3s
retries: 10
oj-judge:
# 自己构建的判题沙箱(go-judge + gcc / python,见 docker/judge/)。
# ⚠️ 这个 tag 不在任何 registry 上:本机 docker/judge/build.sh 构建,
# 服务器/机房用 docker load 装。镜像不在本地时 compose 会去 pull 然后报找不到。
image: oj2-judge-3
container_name: oj-judge
restart: always
# go-judge 要自己建 namespace 和 cgroup 来关学生程序,只能给特权,理由见 docker/judge/README.md
privileged: true
# go-judge 的文件存储(编译产物、程序输出)放这里,见 docker/judge/Dockerfile 的 -dir
tmpfs:
- /tmp
volumes:
- ${DATA_DIR:-../data}/backend/test_case:/test_case:ro
environment:
ES_AUTH_TOKEN: ${OJ2_JUDGE_TOKEN:?请在 env 文件里设置 OJ2_JUDGE_TOKEN}
# 机房机器内存宽裕,判题给足
mem_limit: 2g
oj-api:
build: &build
context: ..
dockerfile: docker/Dockerfile
target: api
# 产物来源,见 docker/Dockerfile 顶部。默认 build(服务器自己编);
# CI 传 ARTIFACTS=prebuilt 用 runner 编好的产物。
args:
ARTIFACTS: ${ARTIFACTS:-build}
image: oj2-api:latest
container_name: oj-api
restart: always
depends_on:
oj-redis:
condition: service_healthy
volumes:
- ${DATA_DIR:-../data}/backend:/data
environment: &api-env
# 库在服务器上,走公网。DB_HOST 默认值就是服务器地址,换机器改 env 文件
DATABASE_URL: postgres://onlinejudge:${POSTGRES_PASSWORD:?}@${DB_HOST:-150.158.29.156}:${DB_PORT:-5445}/onlinejudge
REDIS_URL: redis://oj-redis:6379
JUDGE_SERVER_URL: http://oj-judge:8080
JUDGE_SERVER_TOKEN: ${OJ2_JUDGE_TOKEN:?}
# 后台「判题机列表」里这一台的名字(worker 替判题机记心跳,见 judge/heartbeat.ts)
JUDGE_SERVER_NAME: 机房
JUDGE_CONCURRENCY: ${JUDGE_CONCURRENCY:-4}
TRIAL_CONCURRENCY: ${TRIAL_CONCURRENCY:-4}
AI_KEY: ${AI_KEY:-}
AI_HINT_DIAGNOSE: ${AI_HINT_DIAGNOSE:-}
healthcheck:
test: ["CMD", "oj2-api", "healthcheck"]
interval: 30s
timeout: 3s
retries: 3
start_period: 10s
mem_limit: 4g
oj-worker:
build: *build
image: oj2-api:latest
container_name: oj-worker
restart: always
depends_on:
- oj-api
volumes:
- ${DATA_DIR:-../data}/backend:/data
environment: *api-env
command: ["oj2-api", "worker"]
mem_limit: 2g
oj-web:
build:
context: ..
dockerfile: docker/Dockerfile
target: web
args:
ARTIFACTS: ${ARTIFACTS:-build}
image: oj2-web:latest
container_name: oj-web
restart: always
depends_on:
- oj-api
volumes:
- ${DATA_DIR:-../data}/backend/log:/data/log
ports:
- "${WEB_PORT:-81}:8000"
mem_limit: 256m