restic(esh-vm-db): a backup that exited 0 nightly while keeping an April dump

Work by a parallel session on 2026-09-12; committed here with the rest of the
day's changes. Rationale in persistent-memory.d/2026-09-12-esh-vm-db-restic-repair.md.

The only visible symptom was a systemd-failed unit from a Sep 6 repository
network timeout after boot. The real fault was quieter and much worse: the
pre-backup hook logged failures as WARN and returned zero, so pg_dumpall could
fail every single night -- it used TCP localhost and wanted a password nobody
supplied -- while restic dutifully backed up the stale April 23 dump still
sitting in the staging directory and reported success. Mongo was fine, which
is part of why it went unnoticed.

Postgres now dumps over the /var/run/postgresql socket with peer auth and -w,
and both database failures now fail the backup rather than masking it, while
still preserving any prior per-DB dump rather than truncating to nothing. An
ERRORS counter replaces the warn-and-continue path, and the staging directory
is overridable via RESTIC_STAGE_DIR so the new test can exercise it.

Adds backup.contract.md, retry.conf and test_pre_backup.py -- three red-green
regression tests covering the failure modes above. systemd drop-ins on both
jobs add network-online ordering plus Restart=on-failure with a 5m delay and a
3-per-hour limit, which addresses the original boot-timeout symptom.

Verified against a real run: snapshot bc5eeaff at 07:01 PDT with a fresh 3.46MB
PG dump, retrieved from the repository with decompression and completion marker
checked (not a full restore). Repository check passed, 99 snapshots. The old
hook and stale dump are preserved root-only at /var/lib/restic/repair-20260912.

configs/restic/esh-vm-db/pre-backup.sh

@@ -10,29 +10,32 @@
 # raw volume capture isn't in the restic source list so these dumps
 # are the ONLY way restic sees DB data.
 #
-# Errors in individual blocks log a WARN; whole script doesn't abort.
+# Required dump failures return nonzero so restic cannot report stale DB data
+# as a successful fresh backup. Previous successful dumps are preserved.
 
 set -euo pipefail
+ERRORS=0
 
-STAGE=/var/lib/restic/stage
-install -d -o root -g root -m 0700 "$STAGE"
+STAGE=${RESTIC_STAGE_DIR:-/var/lib/restic/stage}
+install -d -m 0700 "$STAGE"
 
 log() { printf '%s pre-backup(esh-vm-db): %s\n' "$(date -Is)" "$*"; }
 warn() { printf '%s pre-backup(esh-vm-db): WARN: %s\n' "$(date -Is)" "$*" >&2; }
+error() { ERRORS=$((ERRORS + 1)); warn "$*"; }
 
 # ---- Postgres ----------------------------------------------------------
 PG_DUMP="$STAGE/pg_dumpall.sql.gz"
-if sudo -u postgres pg_isready -h localhost -p 5432 > /dev/null 2>&1; then
+if sudo -u postgres pg_isready -h /var/run/postgresql -p 5432 > /dev/null 2>&1; then
   log "pg_dumpall starting → $PG_DUMP"
-  if sudo -u postgres pg_dumpall -h localhost -p 5432 | gzip > "$PG_DUMP.tmp"; then
+  if (cd /; sudo -u postgres pg_dumpall -w -h /var/run/postgresql -p 5432) | gzip > "$PG_DUMP.tmp"; then
     mv "$PG_DUMP.tmp" "$PG_DUMP"
     log "pg_dumpall done ($(du -h "$PG_DUMP" | cut -f1))"
   else
-    warn "pg_dumpall failed (exit $?); keeping previous dump if any"
+    error "pg_dumpall failed (exit $?); keeping previous dump if any"
     rm -f "$PG_DUMP.tmp"
   fi
 else
-  warn "postgres not ready on :5432 — skipping pg_dumpall"
+  error "postgres not ready on :5432 — skipping pg_dumpall"
 fi
 
 # ---- MongoDB -----------------------------------------------------------
@@ -45,15 +48,19 @@ if mongosh --quiet --eval 'db.adminCommand({ping: 1}).ok' | grep -q '^1$'; then
     mv "$MONGO_DIR.tmp" "$MONGO_DIR"
     log "mongodump done ($(du -sh "$MONGO_DIR" | cut -f1))"
   else
-    warn "mongodump failed (exit $?); keeping previous dump if any"
+    error "mongodump failed (exit $?); keeping previous dump if any"
     rm -rf "$MONGO_DIR.tmp"
   fi
 else
-  warn "mongo not reachable via mongosh — skipping mongodump"
+  error "mongo not reachable via mongosh — skipping mongodump"
 fi
 
 # ---- Retention on stage dir --------------------------------------------
 # restic dedupes identical dumps at the chunk level, so we can safely keep
 # overwriting the same files. No explicit rotation needed here.
 
+if [ "$ERRORS" -ne 0 ]; then
+  log "pre-backup FAILED: $ERRORS required dump(s) failed"
+  exit 1
+fi
 log "pre-backup complete"