feat(rebuild): shard the weekly rebuild and make each shard resumable (#163)

## Problem

`weekly-rebuild-missing` runs one job per `<os>-<arch>` and walks that slot's list serially in a single `R -q -e` argument.
That was cheap while every source fallback was skipped as "already built".
Since bincraft #105/#106/#107 and #159 the gate works, and the lists are large: 8 917 source-served records on `amd64/alpine324`, 15 023 on `amd64/resolute`.

Pipeline 10910 (`weekly_rebuild_missing:alpine-324-amd64`) ran for two days, reached `[8692/23885] cholera`, and was killed there.

Two failures follow from that shape:

- **No parallelism.** The work is embarrassingly parallel across packages; one job does all of it.
- **No resumability and no clean stopping point.** The loop ends only by exhausting the list, so the only way to stop it is a kill. A restart re-walks from the first entry, paying a CRAN version resolution and an S3 `HEAD` per package before reaching new work. And a kill matches neither `success` nor `failure`, so the `Purge CDN cache` step never ran: the ~4 600 binaries 10910 did publish stayed hidden behind stale edge copies.

## What this changes

**Three shards per slot.** Each of the 18 `OS`/`ARCH` rows gains `SPLIT_INTO`/`SPLIT_INDEX`, mirroring `build-all-versions.yaml`. Cron and manual routing are unchanged: both filters already match on `${OS}-${ARCH}`, so they now match all three shards of a slot.

**`local/rebuild-missing.R`** replaces the ~1 500-character inline one-liner. The slice is interleaved rather than contiguous, because the list is alphabetical and cost clusters by name (`Rcpp*`, `Bioc*`, `rstan*`).

**Resume by re-deriving state from the bucket.** One `s3_dir_info()` listing gives ETags for the slot; a package is outstanding iff its object's ETag equals CRAN's published `MD5sum`, i.e. it is still byte-identical to CRAN's source. That is `check_s3_root_package()` evaluated in bulk. No progress file, no volume, no DB cursor, and correct when a sibling shard or a `process-updates` run completes something concurrently.

It reads ETags rather than the index's `Built` field the way `packages-to-build.R` does, because the index is no longer rewritten until the dependent pipeline runs and so cannot reflect the current run's progress.

Unknown always means "already a binary", never "rebuild it": a multipart ETag, an unreadable CRAN index or an empty listing can never mass-schedule work.

**A 20 h wall-clock budget** per shard. It exits 0, so the re-index and purge always fire and the remainder is picked up next run with no bookkeeping.

**`.crow/weekly-rebuild-reindex.yaml`** takes over re-indexing and the purge, with `depends_on: [weekly-rebuild-missing]` and `runs_on: [success, failure]`. Three shards writing one slot's `PACKAGES` concurrently would race: `update_PACKAGES()` lists the live bucket, so an early lister that uploads last publishes an index missing its siblings' work.

## Verification

`crow lint .crow/` passes on all 11 pipelines. `prek run` passes.

19 assertions in `local/tests/test-rebuild-missing.R`, 0 failures, covering the partition (disjoint, covering, deterministic, short lists, out-of-range index) and the outstanding filter (source ETag kept, binary ETag dropped, absent object kept, multipart and missing-from-CRAN treated as built).

One of those tests caught a real bug before it shipped: an empty ETag table indexed to zero length rather than to `NA`, which recycled the result away and reported "nothing to build" — the dangerous direction. Fixed with an explicit `lookup()`.

The filter run against the live `amd64/alpine324` index, using its `MD5sum` column as the ETag (established to match the objects):

```
index packages:               24343
outstanding (filter):          8950
no Built stamp:                8917
filter vs no-Built agreement:  8917 of 8917
outstanding but stamped Built:   33 (version drift vs CRAN)
shard sizes: 2984/2983/2983 (sum 8950, unique 8950)
```

It reproduces the source-served set exactly. The extra 33 are packages whose slot version differs from CRAN's current one, so no object exists at the CRAN version key: correctly outstanding.

## Notes for review

- The 20 h budget is a chosen default, exposed as `REBUILD_BUDGET_HOURS` in the pipeline.
- `depends_on` is file-level, not row-level, so on a full cron run no slot is re-indexed until the slowest of all 54 jobs finishes. The budget bounds that at roughly a day.
- An explicit cancel still skips the re-index. Recovery is to trigger `weekly-rebuild-reindex` on its own.
- The purge runs on every re-index row rather than one designated slot: a cron fires only its own slot's row, so gating on a named slot would leave every other slot unpurged.
- Out of scope: `build-all-versions` still cannot rebuild source fallbacks, because `local/build-all.R:113-122` drops every version with any `single_builds` row, which is precisely the source-fallback set.

Design: `specs/2026-08-12-shard-weekly-rebuild-design.md`
Reviewed-on: #163

.crow/weekly-rebuild-missing.yaml

@@ -1,12 +1,21 @@
 # Consolidated weekly-rebuild-missing pipeline (all platforms, both arches).
-# One matrix row per OS/arch replaces the former per-platform files.
+# Three matrix rows per OS/arch, one per shard of that slot's rebuild list.
 # Routing is preserved 1:1:
 #   - cron:   each existing `weekly-rebuild-missing-<os>-<arch>` cron fires only
-#             its matching matrix row (via the per-row `cron:` name filter).
+#             its matching matrix rows (via the per-row `cron:` name filter),
+#             which is now all three shards of that slot.
 #   - manual: `weekly_rebuild_missing` dropdown, default "all" (matches the
 #             previous bare manual trigger that ran every os/arch); pick a
 #             single <os>-<arch> to run just one.
 # Arch placement is handled by the group label (rpkgs-amd64, rpkgs-arm64).
+#
+# The shard picks up its own slice and re-derives what is still outstanding
+# from the bucket, so a restart resumes rather than replaying; see
+# local/rebuild-missing.R.
+#
+# Re-indexing and the CDN purge deliberately do NOT live here. Three shards
+# writing one slot's PACKAGES concurrently would race, so they moved to
+# .crow/weekly-rebuild-reindex.yaml, which depends on this pipeline.
 variables:
   # Gates this pipeline. A manual pipeline creation instantiates every file in
   # .crow/, and a declared default is applied even when the run never passed
@@ -53,74 +62,326 @@ matrix:
       ARCH: amd64
       R_VERSION: 4.5.3
       IMG: alpine:3.22
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: alpine-322
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.22
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: alpine-322
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.22
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: alpine-322
       ARCH: arm64
       R_VERSION: 4.5.3
       IMG: alpine:3.22
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: alpine-322
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.22
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: alpine-322
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.22
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: alpine-323
       ARCH: amd64
       R_VERSION: 4.5.3
       IMG: alpine:3.23
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: alpine-323
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.23
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: alpine-323
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.23
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: alpine-323
       ARCH: arm64
       R_VERSION: 4.5.3
       IMG: alpine:3.23
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: alpine-323
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.23
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: alpine-323
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.23
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: alpine-324
       ARCH: amd64
       R_VERSION: 4.5.3
       IMG: alpine:3.24
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: alpine-324
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.24
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: alpine-324
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.24
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: alpine-324
       ARCH: arm64
       R_VERSION: 4.5.3
       IMG: alpine:3.24
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: alpine-324
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.24
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: alpine-324
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: alpine:3.24
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: redhat-8
       ARCH: amd64
       R_VERSION: 4.4.3
       IMG: redhat:8
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: redhat-8
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: redhat:8
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: redhat-8
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: redhat:8
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: redhat-8
       ARCH: arm64
       R_VERSION: 4.4.3
       IMG: redhat:8
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: redhat-8
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: redhat:8
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: redhat-8
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: redhat:8
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: redhat-9
       ARCH: amd64
       R_VERSION: 4.4.3
       IMG: redhat:9
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: redhat-9
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: redhat:9
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: redhat-9
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: redhat:9
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: redhat-9
       ARCH: arm64
       R_VERSION: 4.4.3
       IMG: redhat:9
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: redhat-9
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: redhat:9
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: redhat-9
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: redhat:9
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: redhat-10
       ARCH: amd64
       R_VERSION: 4.5.3
       IMG: redhat:10
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: redhat-10
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: redhat:10
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: redhat-10
+      ARCH: amd64
+      R_VERSION: 4.5.3
+      IMG: redhat:10
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: redhat-10
       ARCH: arm64
       R_VERSION: 4.5.3
       IMG: redhat:10
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: redhat-10
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: redhat:10
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: redhat-10
+      ARCH: arm64
+      R_VERSION: 4.5.3
+      IMG: redhat:10
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: ubuntu-2204
       ARCH: amd64
       R_VERSION: 4.4.3
       IMG: ubuntu:jammy
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: ubuntu-2204
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:jammy
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: ubuntu-2204
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:jammy
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: ubuntu-2204
       ARCH: arm64
       R_VERSION: 4.4.3
       IMG: ubuntu:jammy
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: ubuntu-2204
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:jammy
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: ubuntu-2204
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:jammy
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: ubuntu-2404
       ARCH: amd64
       R_VERSION: 4.4.3
       IMG: ubuntu:noble
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: ubuntu-2404
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:noble
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: ubuntu-2404
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:noble
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: ubuntu-2404
       ARCH: arm64
       R_VERSION: 4.4.3
       IMG: ubuntu:noble
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: ubuntu-2404
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:noble
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: ubuntu-2404
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:noble
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: ubuntu-2604
       ARCH: amd64
       R_VERSION: 4.4.3
       IMG: ubuntu:resolute
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: ubuntu-2604
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:resolute
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: ubuntu-2604
+      ARCH: amd64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:resolute
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
     - OS: ubuntu-2604
       ARCH: arm64
       R_VERSION: 4.4.3
       IMG: ubuntu:resolute
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 1
+    - OS: ubuntu-2604
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:resolute
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 2
+    - OS: ubuntu-2604
+      ARCH: arm64
+      R_VERSION: 4.4.3
+      IMG: ubuntu:resolute
+      SPLIT_INTO: 3
+      SPLIT_INDEX: 3
 
 steps:
   - name: 'Rebuild missing binaries'
@@ -153,6 +414,12 @@ steps:
       PLATFORM: ${OS}
       ARCH: ${ARCH}
       NCPUS: 2
+      SPLIT_INTO: ${SPLIT_INTO}
+      SPLIT_INDEX: ${SPLIT_INDEX}
+      # Wall clock after which the shard stops cleanly instead of having to be
+      # killed. A kill matches neither `success` nor `failure`, so it would skip
+      # the dependent re-index and leave rebuilt binaries behind a stale edge.
+      REBUILD_BUDGET_HOURS: 20
     commands:
       - git clone -q https://pat-s:$$REPO_RO_TOKEN@git.devxy.io/devxy/build-cran-binaries.git .
       - mkdir -p /mnt/cache/uvr/cache /mnt/cache/uvr/packages /mnt/cache/R-pkgs /mnt/cache/ccache /mnt/cache/packages
@@ -162,18 +429,7 @@ steps:
       - 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
       - UVR_R_BIN=/opt/R/$R_VERSION/bin/R local/uvr-install.sh httr2
       - /opt/R/$R_VERSION/bin/R -q -e 'source("local/fetch-rebuild-packages-from-issue.R")'
-      - $XVFB $XVFB_ARGS -- /opt/R/$R_VERSION/bin/R -q -e "sink(stdout(), type = 'message'); options(crayon.enabled = TRUE, Ncpus = $NCPUS, future.globals.onReference = NULL); pkgs <- readLines('/tmp/rebuild_pkgs.txt'); if (length(pkgs) == 0) { cat('Nothing to rebuild\n'); q('no') }; excluded <- jsonlite::fromJSON('local/excluded-packages.json')[['package']]; pkgs <- setdiff(pkgs, excluded); cat(sprintf('Rebuilding %d packages\n', length(pkgs))); n <- length(pkgs); for (i in seq_along(pkgs)) { x <- pkgs[i]; cat(sprintf('[%d/%d] %s\n', i, n, x)); tryCatch(bincraft::build_binary_package(x, tag_limit = 1L, patches = 'local/patches', 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'), metadata_db_host = 'r-binaries.devxy.io', metadata_db_name = 'build_metadata', metadata_db_table = 'single_builds', metadata_db_user = 'rpkgs', metadata_db_password = Sys.getenv('PGPASS'), metadata_db_sslmode = 'require', metadata_db_port = 15432, archive = TRUE, upload = TRUE, store_build_metadata = TRUE), error = function(e) cat(sprintf('ERROR building %s - %s\n', x, conditionMessage(e)))) }" 2>&1
-      # A rebuild replaces objects in place, so the slot's index still advertises
-      # the old MD5 and, for anything that had been served from source, no Built
-      # stamp. Re-index here rather than waiting for the next process-updates
-      # run, or the rebuilt binaries stay invisible to clients until then.
-      # The codename is detected from the image's /etc/os-release.
-      - /opt/R/$R_VERSION/bin/R -q -e 'library(bincraft); upload_package_index(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 "library(bincraft); upload_package_index(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
+      - $XVFB $XVFB_ARGS -n $SPLIT_INDEX -- /opt/R/$R_VERSION/bin/Rscript local/rebuild-missing.R $SPLIT_INTO $SPLIT_INDEX $REBUILD_BUDGET_HOURS 2>&1
     backend_options:
       docker:
         resources:
@@ -183,25 +439,3 @@ steps:
           limits:
             memory: 18Gi
             cpu: 3000m
-
-  - name: Purge CDN cache
-    image: reg.devxy.io/docker.io/library/alpine:3.24
-    environment:
-      OTEL_R_TRACES_EXPORTER: none
-      OTEL_R_LOGS_EXPORTER: none
-      OTEL_R_METRICS_EXPORTER: none
-      BUNNYNET_API_KEY:
-        from_secret: BUNNYNET_API_KEY
-      REPO_RO_TOKEN:
-        from_secret: REPO_RO_TOKEN
-      # All hostnames on the zone share this id, so one purge covers
-      # cran.devxy.io, cran.allianceswisspass.devxy.io and cran.rpkgs.com.
-      BUNNY_PULLZONE: '3857050'
-    commands:
-      - apk add --no-cache -q bash curl git
-      - git clone -q https://pat-s:$$REPO_RO_TOKEN@git.devxy.io/devxy/build-cran-binaries.git .
-      - bash scripts/purge_cdn_zone.sh "$BUNNYNET_API_KEY" "$BUNNY_PULLZONE"
-    # A rebuild that died part-way still replaced objects, and those are exactly
-    # the ones a stale edge would keep hiding, so purge either way.
-    when:
-      - status: [success, failure]