build-cran-binaries/.crow/build-all-versions.yaml
pat-s a4b274f281
Some checks failed
ci/crow/cron/process-updates/7 Pipeline was successful
ci/crow/cron/process-updates/3 Pipeline was successful
ci/crow/cron/process-updates/9 Pipeline was canceled
ci/crow/cron/process-updates/4 Pipeline was successful
ci/crow/manual/build-all-versions-install-deps/2 Pipeline was successful
ci/crow/cron/process-updates/13 Pipeline was canceled
ci/crow/cron/process-updates/8 Pipeline failed
ci/crow/cron/process-updates/10 Pipeline was successful
ci/crow/cron/process-updates/17 Pipeline was canceled
ci/crow/cron/process-updates/11 Pipeline was successful
ci/crow/cron/process-updates/5 Pipeline failed
ci/crow/manual/build-all-versions-install-deps/1 Pipeline was successful
ci/crow/cron/process-updates/15 Pipeline failed
ci/crow/cron/process-updates/1 Pipeline was successful
ci/crow/cron/process-updates/14 Pipeline was successful
ci/crow/manual/build-all-versions/2 Pipeline was successful
ci/crow/cron/process-updates/16 Pipeline was successful
ci/crow/manual/build-all-versions/3 Pipeline failed
ci/crow/manual/build-all-versions/1 Pipeline failed
ci/crow/cron/process-updates/18 Pipeline was successful
ci/crow/manual/build-all-versions/6 Pipeline failed
ci/crow/manual/build-all-versions/5 Pipeline was canceled
ci/crow/manual/build-all-versions/8 Pipeline was canceled
ci/crow/cron/process-updates/12 Pipeline was canceled
ci/crow/manual/build-all-versions/7 Pipeline was canceled
ci/crow/cron/process-updates/6 Pipeline failed
ci/crow/cron/process-updates/2 Pipeline failed
ci/crow/manual/build-all-versions/4 Pipeline was canceled
fix(ci): bound pkgcache _metadata growth to stop macmini disk-fill (#110)
## Problem

The arm64 `build-all` pipeline fills the macmini (gaia) host disk despite an 8h prune.
Root cause is not images or job volumes: it is the persistent dep-cache volume, specifically `pkgcache/R/pkgcache/_metadata`, which grew to ~165 GB.
`{pkgcache}` mints a new content hash for the "patched" binaries repo on every PACKAGES change, so each per-package build writes a fresh ~70 MB `pkgs-<hash>.rds` (+ `patched-<hash>/`) that is never evicted (2407 snapshots observed).
When the disk hits 100% OrbStack stops and the on-host prune can no longer connect to the daemon, so it never self-heals.

## Change (Workstream A of the disk-fill fix)

- Add `trim_pkgcache_metadata()` to `local/r-minor-helpers.R`: keeps the newest `keep` (default 20) `patched-*`/`pkgs-*.rds` entries under `_metadata`, deleting only entries older than `min_age_secs` (default 600s) so it never races the up-to-4 concurrent split-jobs sharing the volume.
Preserves `pkg/` downloads and the stable CRAN/BioC/INLA repo dirs.
No-op when `R_PKG_CACHE_DIR` is empty (amd64) or `_metadata` is absent (first run).
- Call it every 25 packages inside the build loop in `local/build-all.R`.
- Add a defensive start-of-run cleanup of `_metadata/patched-*` + `pkgs-*.rds` to the two workflows that mount the persistent volume (`build-all-versions.yaml`, `build-all-versions-install-deps.yaml`).

Only these paths are touched; `process-updates.yaml`/`weekly-rebuild-missing.yaml` (no persistent volume) are unchanged.

Follow-ups (separate workstreams): on-host self-healing prune watcher + OrbStack disk cap (ansible), and Prometheus/Grafana alerting (k8s-talos).
Upstream: bincraft patched-repo hash churn is the true source fix.

New unit tests (6) for the helper; full suite 24/24 green.

Reviewed-on: #110
2026-07-13 09:31:29 +00:00

189 lines
7.3 KiB
YAML

### Manual full rebuild of all CRAN binaries for one OS/arch.
# Set the variables below in the manual-run form (or via --var on the CLI).
# `target_arch` routes the matrix; OS/OS_VERSION/R_VERSION select the platform
# image and cache volume. Placement is via the group label (rpkgs-amd64/rpkgs-arm64).
# Skip list lives in local/excluded-packages.json (read by local/build-all.R).
variables:
target_arch:
description: 'Architecture to build.'
options:
- amd64
- arm64
default: amd64
OS:
description: "Base OS image name."
options:
- alpine
- redhat
- ubuntu
default: alpine
OS_VERSION:
description: "OS image tag. Must match OS (alpine: 3.24; redhat: 8/9/10; ubuntu: jammy/noble)."
options:
- "3.22"
- "3.23"
- "3.24"
- "8"
- "9"
- "10"
- "jammy"
- "noble"
- "resolute"
default: "3.24"
R_VERSION:
description: 'Primary R version under /opt/R.'
options:
- 4.5.3
- 4.4.3
default: 4.5.3
when:
- event: manual
evaluate: 'target_arch == "${ARCH}"'
skip_clone: true
labels:
platform: linux/${ARCH}
group: rpkgs-${ARCH}
matrix:
include:
- ARCH: amd64
R_PKG_CACHE_DIR: ''
SPLIT_INTO: 4
SPLIT_INDEX: 1
- ARCH: amd64
R_PKG_CACHE_DIR: ''
SPLIT_INTO: 4
SPLIT_INDEX: 2
- ARCH: amd64
R_PKG_CACHE_DIR: ''
SPLIT_INTO: 4
SPLIT_INDEX: 3
- ARCH: amd64
R_PKG_CACHE_DIR: ''
SPLIT_INTO: 4
SPLIT_INDEX: 4
- ARCH: arm64
R_PKG_CACHE_DIR: /mnt/cache/pkgcache
SPLIT_INTO: 4
SPLIT_INDEX: 1
- ARCH: arm64
R_PKG_CACHE_DIR: /mnt/cache/pkgcache
SPLIT_INTO: 4
SPLIT_INDEX: 2
- ARCH: arm64
R_PKG_CACHE_DIR: /mnt/cache/pkgcache
SPLIT_INTO: 4
SPLIT_INDEX: 3
- ARCH: arm64
R_PKG_CACHE_DIR: /mnt/cache/pkgcache
SPLIT_INTO: 4
SPLIT_INDEX: 4
depends_on:
- build-all-versions-install-deps
steps:
- name: 'Build binaries'
image: 'reg.devxy.io/rpkgs/build-env-${OS}:${OS_VERSION}'
pull: true
environment:
OTEL_R_TRACES_EXPORTER: none
OTEL_R_LOGS_EXPORTER: none
OTEL_R_METRICS_EXPORTER: none
RED_HAT_DEV_PW:
from_secret: RED_HAT_DEV_PW
B2_S3_ACCESS_KEY:
from_secret: B2_S3_ACCESS_KEY
B2_S3_SECRET_KEY:
from_secret: B2_S3_SECRET_KEY
PGPASS:
from_secret: PGPASS
REPO_RO_TOKEN:
from_secret: REPO_RO_TOKEN
GITHUB_PAT:
from_secret: GITHUB_PAT
# normal env vars
GIT_USER: pat-s
# set the location of the 'pkgcache' cache dir which persists the R package dependencies needed to install the packages themselves
R_PKG_CACHE_DIR: ${R_PKG_CACHE_DIR}
R_LIBS_USER: /mnt/cache/R-pkgs
CCACHE_DIR: /mnt/cache/ccache
NCPUS: 2
volumes:
- ${ARCH}-binaries-r-dep-cache-${OS}-${OS_VERSION//./}:/mnt/cache
commands:
- git clone -q https://pat-s:$$REPO_RO_TOKEN@git.devxy.io/devxy/build-cran-binaries.git .
# Clear churny pkgcache metadata left by a prior crashed run (the "patched"
# repo mints a new hash per PACKAGES change -> unbounded pkgs-*.rds/patched-*).
# Keep pkg/ downloads and the stable CRAN/BioC/INLA repo dirs. Within-run
# growth is bounded separately by trim_pkgcache_metadata() in build-all.R.
- rm -rf /mnt/cache/pkgcache/R/pkgcache/_metadata/patched-* /mnt/cache/pkgcache/R/pkgcache/_metadata/pkgs-*.rds || true
- mkdir -p /mnt/cache/pkgcache /mnt/cache/R-pkgs /mnt/cache/ccache /mnt/cache/packages
# The primary pass must not rely on build-all-versions-install-deps having
# run on *this* agent: depends_on only orders the steps, but the cache
# volume is per-agent, so a job landing on an agent where install-deps did
# not run would otherwise use a stale bincraft (which resolves `platform`
# to a zero-length value and breaks every metadata query and the sysdeps
# install). Pin bincraft here, exactly like the R-minor pass below.
- rm -rf /mnt/cache/R-pkgs/00LOCK-*
- /opt/R/$R_VERSION/bin/Rscript local/install-bincraft.R
- XVFB=$(command -v xwfb-run 2>/dev/null || command -v xvfb-run); XVFB_ARGS=""; if command -v xwfb-run >/dev/null 2>&1; then dnf install -y -q weston 2>/dev/null; XVFB_ARGS="-c weston"; fi
- $XVFB $XVFB_ARGS -n $SPLIT_INDEX -- /opt/R/$R_VERSION/bin/Rscript local/build-all.R $SPLIT_INTO $SPLIT_INDEX $NCPUS 2>&1
- |
PRIMARY_MINOR=$(echo "$R_VERSION" | cut -d. -f1-2)
for RBIN in /opt/R/[0-9]*/bin/R; do
RV=$(basename "$(dirname "$(dirname "$RBIN")")")
RMINOR=$(echo "$RV" | cut -d. -f1-2)
[ "$RMINOR" = "$PRIMARY_MINOR" ] && continue
echo "=== R-minor-sensitive pass under R $RV ==="
LIB="/mnt/cache/R-pkgs-$RMINOR"
mkdir -p "$LIB"
R_LIBS_USER="$LIB" "$(dirname "$RBIN")/Rscript" local/install-bincraft.R || true
R_LIBS_USER="$LIB" $XVFB $XVFB_ARGS -n $SPLIT_INDEX -- "$(dirname "$RBIN")/Rscript" local/build-all.R --sensitive-only $SPLIT_INTO $SPLIT_INDEX $NCPUS 2>&1 || true
done
# archive missed packages; first arg is the codename (e.g. "alpine324"),
# derived via bincraft like the upload step, not paste(OS, OS_VERSION).
- /opt/R/$R_VERSION/bin/R -q -e "bincraft::process_unarchived_pkgs(bincraft::set_codename(NULL), Sys.getenv('ARCH'), s3_endpoint = 'https://s3.eu-central-003.backblazeb2.com', s3_region = 'eu-central-003', s3_bucket = 'devxy-rpkgs-binaries', s3_access_key_id = Sys.getenv('B2_S3_ACCESS_KEY'), s3_secret_access_key = Sys.getenv('B2_S3_SECRET_KEY'), workers = $NCPUS)"
backend_options:
docker:
resources:
requests:
memory: 5Gi
cpu: 1000m
limits:
memory: 20Gi
cpu: 2000m
- name: 'Upload package indexes'
image: 'reg.devxy.io/rpkgs/build-env-${OS}:${OS_VERSION}'
pull: true
environment:
OTEL_R_TRACES_EXPORTER: none
OTEL_R_LOGS_EXPORTER: none
OTEL_R_METRICS_EXPORTER: none
B2_S3_ACCESS_KEY:
from_secret: B2_S3_ACCESS_KEY
B2_S3_SECRET_KEY:
from_secret: B2_S3_SECRET_KEY
R_LIBS_USER: /mnt/cache/R-pkgs
volumes:
- ${ARCH}-binaries-r-dep-cache-${OS}-${OS_VERSION//./}:/mnt/cache
commands:
- |
CODENAME=$(/opt/R/$R_VERSION/bin/Rscript -e "cat(bincraft::set_codename(NULL))")
/opt/R/$R_VERSION/bin/R -q -e "bincraft::upload_package_index(codename = '$CODENAME', s3_endpoint = 'https://s3.eu-central-003.backblazeb2.com', s3_region = 'eu-central-003', s3_bucket = 'devxy-rpkgs-binaries', s3_access_key_id = Sys.getenv('B2_S3_ACCESS_KEY'), s3_secret_access_key = Sys.getenv('B2_S3_SECRET_KEY'))"
for RBIN in /opt/R/[0-9]*/bin/R; do
RMINOR=$(basename "$(dirname "$(dirname "$RBIN")")" | cut -d. -f1-2)
/opt/R/$R_VERSION/bin/R -q -e "bincraft::upload_package_index(codename = '$CODENAME', r_minor = '$RMINOR', s3_endpoint = 'https://s3.eu-central-003.backblazeb2.com', s3_region = 'eu-central-003', s3_bucket = 'devxy-rpkgs-binaries', s3_access_key_id = Sys.getenv('B2_S3_ACCESS_KEY'), s3_secret_access_key = Sys.getenv('B2_S3_SECRET_KEY'))" || true
done
backend_options:
docker:
resources:
requests:
memory: 5Gi
cpu: 1000m
limits:
memory: 20Gi
cpu: 2000m