diff --git a/docs/architecture/backup-dr.md b/docs/architecture/backup-dr.md
index 93c292dc..df30dea2 100644
--- a/docs/architecture/backup-dr.md
+++ b/docs/architecture/backup-dr.md
@@ -365,6 +365,7 @@ The hand-managed Linux VMs are **intentionally not in Terraform** (telmate/bpg p
 **Mode**: `vzdump --mode snapshot` — live, no downtime. devvm has the qemu guest agent enabled (`agent: 1`), so the snapshot is **filesystem-consistent** (fs-freeze) rather than merely crash-consistent. Runs `Nice=10` + `IOSchedulingClass=idle` + `--ionice 7` so it never starves etcd on the contended sdc IO domain.
 **Scope**: VMIDs in `VZDUMP_VMIDS` (default `102` = devvm). Add VMIDs there to image other hand-managed VMs.
 **Retention**: `KEEP=3` newest dumps per VMID on sda (`/mnt/backup/vzdump/`); each devvm image is ~35-50 GB zstd.
+**Critical dependency**: `nfs-mirror` MUST keep `--exclude='/vzdump/'`. Its nightly `rsync -rlt --delete /srv/nfs/ → /mnt/backup/` treats any `/mnt/backup` dir with no `/srv/nfs` counterpart as an orphan and deletes it — this silently reaped the first two vzdump images at 02:00 on 2026-06-10 before the exclude was added (same reason `pvc-data`/`pfsense`/`pve-config`/`sqlite-backup` are excluded).
 **Offsite**: deliberately **NOT** appended to the incremental offsite manifest — it never deletes, so daily multi-GB images would accumulate unbounded on Synology. Instead the **monthly offsite-sync full pass (days 1-7)** mirrors all of `/mnt/backup` (including `vzdump/`) to Synology with `--delete`, bounded to local retention. So Copy 2 (sda) refreshes **daily**; Copy 3 (Synology) refreshes **monthly**.
 **Monitoring**: pushes `vzdump_last_run_timestamp` / `vzdump_last_status` / `vzdump_last_success_timestamp` to Pushgateway job `vzdump-backup`. A `VzdumpBackupStale` / `VzdumpBackupFailing` alert in `stacks/monitoring` (mirroring the LVM/pfSense backup alerts) is the recommended next addition.
 **Restore**: on the PVE host, `qmrestore /mnt/backup/vzdump/vzdump-qemu-<vmid>-<ts>.vma.zst <vmid>` — restore to a spare VMID first if the original still exists, then swap disks; or use the PVE UI (add `/mnt/backup` as a dir storage with content=backup → Restore).
diff --git a/scripts/nfs-mirror.sh b/scripts/nfs-mirror.sh
index 2e322ede..882a8a9c 100644
--- a/scripts/nfs-mirror.sh
+++ b/scripts/nfs-mirror.sh
@@ -54,11 +54,14 @@ PUSHGATEWAY="${NFS_MIRROR_PUSHGATEWAY:-http://10.0.20.100:30091}"
 PUSHGATEWAY_JOB=nfs-mirror
 
 EXCLUDES=(
-    # ---- /mnt/backup subtrees owned by daily-backup — leave alone ----
+    # ---- /mnt/backup subtrees owned by OTHER backup jobs — leave alone ----
+    # Without these, the top-level `rsync --delete /srv/nfs/ → /mnt/backup/` below
+    # reaps any /mnt/backup dir that has no /srv/nfs counterpart.
     --exclude='/pvc-data/'
     --exclude='/sqlite-backup/'
     --exclude='/pfsense/'
     --exclude='/pve-config/'
+    --exclude='/vzdump/'       # VM images from vzdump-vms — NOT a /srv/nfs svc (else --delete reaps them nightly)
     --exclude='/lost+found/'
 
     # ---- state files used by other backup jobs ----