LibrePortal/scripts/backup/verify/backup_verify.sh
librelad 036f72d3c2 fix(backup): verify against snapshot restorability, not the live dir
With live dumps + container-side file captures the live app dir intentionally
differs from the snapshot (raw DB dirs and private trees are excluded, replaced
by dumps/captures), so the old source-vs-restored file-count check false-failed.
The scratch restore succeeding already proves restorability (restic hash-checks
every blob); keep a non-empty sanity check instead.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
2026-05-23 18:27:28 +01:00

44 lines
1.5 KiB
Bash

#!/bin/bash
backupVerifySnapshot()
{
local idx="$1"
local snapshot_id="$2"
local app_name="$3"
if [[ -z "$snapshot_id" ]]; then
isNotice "No snapshot id provided — verification skipped"
return 0
fi
local scratch
scratch=$(mktemp -d -t libreportal-verify-XXXXXX)
sudo chown "$docker_install_user":"$docker_install_user" "$scratch"
isNotice "Verifying ${snapshot_id:0:8} via scratch restore at $scratch"
if ! engineRestoreSnapshot "$idx" "$snapshot_id" "$scratch" "$containers_dir$app_name"; then
isError "Verify restore FAILED for $app_name on $(resticLocationName "$idx")"
sudo rm -rf "$scratch"
return 1
fi
# The scratch restore above succeeding is the real proof the snapshot is
# restorable (restic verifies each blob's hash as it restores). We can't
# compare against the live app dir any more — the live path deliberately
# differs from the snapshot (raw DB dirs and private file trees are excluded
# and replaced by dumps/captures under .lp-backup) — so just sanity-check the
# restore produced a non-empty tree.
local restored_count
restored_count=$(sudo find "$scratch$containers_dir$app_name" -type f 2>/dev/null | wc -l)
sudo rm -rf "$scratch"
if [[ "$restored_count" -lt 1 ]]; then
isError "Verify FAILED for $app_name — restored snapshot is empty"
return 1
fi
isSuccessful "Snapshot ${snapshot_id:0:8} verified — restored $restored_count files"
}