fix: report failed backup monitoring when the backup never starts

[MarcWort]

Motivation

A backup that aborts before restic begins because of failed run-before or immediate exit from stdin left the previous result in the prometheus metrics file, so dashboards and alerting kept showing a healthy backup even after a failure.

How it works

Both abort points now report the failure through the existing r.summary(...) plumbing, so the monitoring receivers write the failed status.

The *_bytes/*_files/*_dir gauges are written as 0 on an interrupted backup because restic's completion summary is never printed.

Disclaimer:

I used a selfhosted qwen3.8:27b with opencode while also trying it myself. I would have done the same changes but with somewhat diffrent tests.

Let me know what you think.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.78%. Comparing base (65f4135) to head (14a1e8e).

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #676      +/-   ##
==========================================
+ Coverage   81.76%   81.78%   +0.02%     
==========================================
  Files         161      161              
  Lines       12009    12013       +4     
==========================================
+ Hits         9818     9824       +6     
+ Misses       1692     1689       -3     
- Partials      499      500       +1     

Flag	Coverage Δ
unittests	80.04% <100.00%> (+0.02%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.