Backup and Disaster Recovery · #64
Online Backup Job Report Analyzer
Finds jobs exceeding RPO, failing repeatedly or running long from backup job history.
- Data is processed in your browser and not sent to a server
- Export: CSV
Tool
What does this tool do?
When daily backup emails have hundreds of lines, 'Failed' rows get missed; a job failing every night may have its last good backup days ago. The tool compares the age of each job's last successful (or warning) run with the RPO threshold and flags consecutive failures and runs longer than twice the average.
How to use it
- Export the job history as CSV from your backup software; header needs
job, start, status,endandsize_gbare optional. - Date format
YYYY-MM-DD HH:MMorDD.MM.YYYY HH:MM; status Success/Warning/Failed (TR/DE equivalents accepted). - Enter the RPO threshold in hours; press Calculate.
Fields
- Job history (CSV)
- Data is processed in the browser, not sent.
- RPO threshold (hours)
- Alert if the last good backup is older.
Worked example
| Input | Value |
|---|---|
| Job history (CSV) | |
| RPO threshold (hours) | 24 |
Expected result: FILE has failed for 3 nights, last good backup 71.2 hours ago: alert. VM-ERP's 20 November run took more than twice the average; SQL-LOG is fine.
The reference time is the latest time in the file (21 November 23:41), not the current clock, so older reports are evaluated consistently. Check the error message in the backup software for FILE; three consecutive failures usually mean credentials, quota or target access issues.
Second scenario
| Input | Value |
|---|---|
| Job history (CSV) | |
| RPO threshold (hours) | 24 |
Expected result: A German report (semicolon, DD.MM.YYYY): Exchange failed in the last 2 runs, last good backup 48 hours ago.
Method and assumptions
Jobs are grouped by name; duration statistics use successful and warning runs only. A warning run counts as a good backup (a restore point exists). Age of last good = reference − end of last good run; the reference is the latest time in the file.
Alert: no good run, age > RPO or last ≥ 2 runs failed. 'Long' flag: with at least 3 durations, the longest run exceeds twice the average.
Using the result and troubleshooting
Prioritize alerting jobs and run them manually after fixing the error. Regularly update the RPO/RTO matrix and the 3-2-1-1-0 check.
Limits and common mistakes
- Time zone information is ignored; all times are assumed to be in one zone. Rename headers if your product uses different column names.
Frequently asked questions
Should a Warning job count as successful?
For RPO yes, because a restore point was created; the warning reason (e.g. skipped files, VSS error) still needs investigation.
Related tools and service
- RPO and RTO Gap MatrixCompares target and actual RPO/RTO per system and lists the gaps.
- 3-2-1-1-0 Backup Rule CheckerAudits each system's backup copies against the five conditions of the 3-2-1-1-0 rule.
Related service: Backup and Disaster Recovery