# RESTORE HEADERONLY — results for disputed Jenkins backups (step 4)

Verification date: 2026-09-16. Method: `RESTORE HEADERONLY FROM DISK` via sqlcmd (read-only, reads only the .bak header).
Note: part of the headers weren't returned in a regular batch — the batch was cut off; resolved by wrapping with `BEGIN TRY ... END TRY BEGIN CATCH ... END CATCH`.

## Facts table

| File | DatabaseName (inside) | ServerName | BackupStart | .bak size GB | Compressed | DB mounted? |
|---|---|---|---|---|---|---|
| Jenkins4_STROHFARMUAP01_07062026.bak | **STROHFARMUAP01** | **U22930128\SQL2019** (old instance!) | 2026-07-06 14:50 | 15.46 | 0 | **NO** |
| Jenkins4_UAP_Palmdale_SHR_251112.bak | **2610PAMDALEUAP01** | u22930128\SQL2022 | 2025-11-12 07:03 | 1.65 | 0 | YES (a fresh one taken in Q1) |
| Jenkins4_2430PalmdaleF01_04082026.bak | **2430PalmdaleF01** | u22930128\SQL2022 | 2026-04-08 18:09 | 8.59 | 0 | **NO** (a different one is mounted: 2430PalmdaleAgnosticF01) |
| Jenkins4_2510JMREYNOLDSUAP101_12122025.bak | **2510JMREYNOLDSUAP101** | u22930128\SQL2022 | 2025-12-12 14:07 | 33.68 | 0 | **NO** (JMREYNOLDSUAP01/cfg are mounted) |
| Jenkins4_2430CHILDERSOILUAP101_07252025.bak | **2430CHILDERSOILUAP101** | u22930128\SQL2022 | 2025-07-25 17:22 | 18.18 | 0 | **NO** (2210CHILDERSOILUAP01 is mounted) |
| Jenkins4_2430CHILDERSOILUAP201_08262025.bak | **2430CHILDERSOILUAP201** | u22930128\SQL2022 | 2025-08-26 14:08 | 3.16 (20.0 GB unpacked) | 1 (MS_XPRESS) | **NO** |
| Jenkins4_2430Palmdale01_01122026.bak | **2430Palmdale01** | u22930128\SQL2022 | 2026-01-12 20:37 | 93.35 | 0 | **NO** (2620Palmdale01 is mounted) |
| D:\Backup\Jenkins4_STROHFARMUAP01_08122026.bak | (not read — in D:\Backup) | — | — | 2.03 (by dir) | — | **NO** (STROHFARM family) |

## Key conclusions

1. **STROHFARMUAP01 — from the OLD SQL2019 instance**, DatabaseVersion=904 (SQL 2019), DB created 2024-03-21. Current instance is SQL2022 (DatabaseVersion=957). This is a **migrated/decommissioned database**, no live copy exists → .bak is the only source. Freshest is 08122026 (2.03 GB, in D:\Backup), 07062026 — 15.46 GB. The 7.6x size difference is suspicious: either 0812 is a differential/truncated one, or the DB sharply shrank. Requires checking the backup type (BackupType).

2. **Five families point to NON-mounted databases** — their .bak are unique:
   STROHFARMUAP01, 2430PalmdaleF01, 2510JMREYNOLDSUAP101, 2430CHILDERSOILUAP101, 2430CHILDERSOILUAP201, 2430Palmdale01

3. **UAP_Palmdale_SHR_251112.bak is 2610PAMDALEUAP01** (mounted). Not an orphan by data, but this is its **historical version** from 2025-11-12 — valuable as a "previous version" (what the operator asked for for small active DBs). Cheap (1.65 GB).

4. **File name ≠ database name inside**: `UAP_Palmdale_SHR` contained `2610PAMDALEUAP01`; `Jenkins4_2430PalmdaleF01` contained `2430PalmdaleF01`. Conclusion: without HEADERONLY, classification by name is unreliable — all disputed ones need a header check.

5. **2430Palmdale01 has 2 versions** (01122026 93.35 GB and 01292026 93.37 GB). Version 01292026 is **already exfiltrated in Q2**. Version 01122026 — an earlier point of the same non-mounted DB; 93.35 GB, expensive (~4.5 h).

## Mounted databases (54) — for verification
From master_files.txt, relevant to the Jenkins-backup families:
2210CHILDERSOILUAP01, 2430PalmdaleAgnosticF01, 2430PalmdaleAgnosticV01, 2430ProdDevPalmdaleUAP01,
2610PAMDALEUAP01, 2620Palmdale01, 2710PALMDALEOILUAP01, JMREYNOLDSUAP01, JMREYNOLDSUAPcfg,
PALMDALEOILUAP01 (+ 44 others, including 2610JOHNSONPETROLEUMUAP01, 2430DDENERGYUAP101, CHERRYENERGYUAP01, 2210PIGOTTOILUAP01, etc.)