fmq
b058e66722
物理修复（400 失败点归因，见 docs/failed400_nan_pseudo_convergence_2026_08_17.md）： - runner: 阶段 converged 后复查 fort.7，含 NaN/Inf 即否决（fort.9 全零伪收敛， 曾致 261 点误跳过 nl_direct 回退）；否决阶段不产出种子，阻断污染传播 - runner: nl_tight 回退——仅能量边际失败时以 CHMAX 收紧 10× 从自身模型续迭代， 残差降幅 ~10×；execute_tlusty_stage 抽取供主链与回退共用 - conv_check: fort.14 为 Eddington 通量 Hλ，积分需乘 4π 再比 σTeff⁴ （旧版 ratio 稳定 0.0796=1/4π，全点系统性假阳性）+ 回归测试 - nst_writer: 单行超 80 字符被 TLUSTY 静默截断，IFALI/JALI/TRAD 等从未生效； 按 75 字符自动换行 - seed_finder: Teff 容忍度改含边界 <=，相邻 5000K 档恢复互为种子 + 回归测试 谱线表与网格： - 默认线表 gfVIS99 → gfATO（全波段 18-23000Å），TaskSpec.linelist 支持工作流 级覆盖，节点按需下载（进程互斥锁防并发重复下载 238MB） - sdB_cno Teff 加密至 5000K 步长，432 → 9216 点；tlusty/synspec 静态二进制更新 统计与部署： - grid 汇总改按 tlusty_status/synspec_status 分项计数，新增 tlusty_failed/ synspec_failed/synspec_pending，前端详情页双视图适配 - deploy/fetch_results 支持跳板机 ProxyJump 与 SSH 主连接复用，fetch 新增 --force； - Docker 构建支持 CARGO_MIRROR/USE_MIRRORS 国内镜像参数；移除 tools/ 拷贝 - 新增 tlusty-synspec-test skill 与 6 篇根因分析/验证文档
87 lines
3.0 KiB
Docker
87 lines
3.0 KiB
Docker
# =============================================================================
|
||
# DCTS Server Dockerfile — Multi-stage Build (Server + Frontend Dashboard)
|
||
# =============================================================================
|
||
|
||
# ─── Stage 1: 前端静态资源构建 ────────────────────────────────────────────────
|
||
FROM node:22-alpine AS frontend-builder
|
||
|
||
# 国内（CN）网络构建时传 --build-arg USE_MIRRORS=1 切换 npm 源；默认关闭避免境外构建失败。
|
||
ARG USE_MIRRORS=0
|
||
WORKDIR /app/dashboard
|
||
|
||
RUN if [ "$USE_MIRRORS" = "1" ]; then \
|
||
npm config set registry https://registry.npmmirror.com; \
|
||
fi
|
||
|
||
COPY dashboard/package.json dashboard/package-lock.json* ./
|
||
RUN npm install
|
||
COPY dashboard/ ./
|
||
RUN npm run build
|
||
|
||
# ─── Stage 2: Rust 服务端编译 (Alpine/musl 静态编译) ─────────────────────────
|
||
FROM rust:1.94-alpine AS backend-builder
|
||
|
||
ARG USE_MIRRORS=0
|
||
|
||
RUN if [ "$USE_MIRRORS" = "1" ]; then \
|
||
sed -i 's|dl-cdn.alpinelinux.org|mirrors.aliyun.com|g' /etc/apk/repositories; \
|
||
fi
|
||
|
||
RUN apk add --no-cache musl-dev g++ make cmake pkgconfig sqlite-dev openssl-dev openssl-libs-static
|
||
|
||
WORKDIR /app
|
||
ENV SKIP_DASHBOARD_BUILD=1
|
||
|
||
COPY Cargo.toml Cargo.lock ./
|
||
COPY crates/ ./crates/
|
||
COPY workflows/ ./workflows/
|
||
|
||
COPY --from=frontend-builder /app/dashboard/dist ./dashboard/dist
|
||
|
||
RUN cargo build --release -p server && \
|
||
cp /app/target/release/server /usr/local/bin/dcts-server
|
||
|
||
# ─── Stage 3: 最小化生产运行镜像 ─────────────────────────────────────────────
|
||
FROM alpine:3.21
|
||
|
||
ARG USE_MIRRORS=0
|
||
|
||
RUN if [ "$USE_MIRRORS" = "1" ]; then \
|
||
sed -i 's|dl-cdn.alpinelinux.org|mirrors.aliyun.com|g' /etc/apk/repositories; \
|
||
fi
|
||
|
||
RUN apk add --no-cache ca-certificates tzdata sqlite
|
||
|
||
# 创建非 root 账号管理运行
|
||
RUN addgroup -g 65532 -S dcts && adduser -u 65532 -S dcts -G dcts
|
||
|
||
WORKDIR /app
|
||
|
||
COPY --from=backend-builder /usr/local/bin/dcts-server /app/
|
||
COPY --from=frontend-builder /app/dashboard/dist ./dashboard/dist
|
||
COPY workflows/ ./workflows/
|
||
|
||
RUN mkdir -p /app/data /app/data/logs /app/data/seeds /app/assets && \
|
||
chown -R dcts:dcts /app
|
||
|
||
USER dcts
|
||
|
||
EXPOSE 8090
|
||
|
||
ENV DCTS_PORT=8090
|
||
ENV DCTS_DB_PATH=/app/data/dcts.db
|
||
ENV DCTS_QUEUE_DB_PATH=/app/data/dcts_queue.db
|
||
ENV DCTS_SEEDS_DIR=/app/data/seeds
|
||
ENV DCTS_BACKUP_DIR=/app/data/backups
|
||
ENV DCTS_ASSETS_DIR=/app/assets
|
||
ENV LOG_DIR=/app/data/logs
|
||
|
||
VOLUME ["/app/data"]
|
||
|
||
# 健康检查走不走鉴权的 /healthz（启用 admin token 后 /api/status 返回 401，
|
||
# 会令容器被误判不健康而反复重启）。与 docker-compose.yml 的 healthcheck 保持一致。
|
||
HEALTHCHECK --interval=30s --timeout=10s --start-period=15s --retries=3 \
|
||
CMD wget -q --spider http://127.0.0.1:8090/healthz || exit 1
|
||
|
||
ENTRYPOINT ["/app/dcts-server"]
|