[SOLVED] Server 2019 v1809 unable to take any Cumulative Updates

[wdjenkins]

I'm fairly desperate in looking for fix for multiple Server 2019 systems I have (they were all built from same SCCM deployment and all seem to have the same issue). Recreating these from scratch would lose me 6+ months of work, so fixing the issue in place is critical. Long story short is that on this build of Server 2019 no Cumulative Updates from Microsoft will install (other updates work fine). I've tried using Windows Update, installing from downloaded update, and using the DISM to install the update with CAB/MSU method. All the same results. SFC /scannow reports no corruptions to repair, DISM /ONLINE /CLEANUP-IMAGE /RESTOREHEALTH always reports successful, and SFCFix reports everything is ok. Looking at the CBS logs, I can see this which I assume is the issue:

Repr: CSI meta data corruption found, will commit repair transaction if repair is asked.
Repr: No owner found for component amd64_023da5fb41dc3982e550c8b366276df5_31bf3856ad364e35_10.0.17763.1697_none_09e444ccd45d648f, it might be a 3rd party component [HRESULT = 0x800f090c - CBS_E_COMPONENT_NOT_INSTALLED_BY_CBS]
Not able to get CBS component onwers for component amd64_023da5fb41dc3982e550c8b366276df5_31bf3856ad364e35_10.0.17763.1697_none_09e444ccd45d648f [HRESULT = 0x800f090c - CBS_E_COMPONENT_NOT_INSTALLED_BY_CBS]
Repr: Non CBS installed component found, ignore it since it is not repairble
Repr: CSI Store check completes
Exec:Only CSI metadata corruption found, Process CSI repair.

All CSI corruption was fixed, ensure CorruptionDetectedDuringAcr is clear
Failed to clear CorruptionDetectedDuringAcr store corrupt flag (slow mode trigger). [HRESULT = 0x80070002 - ERROR_FILE_NOT_FOUND]

 

[wdjenkins]

Windows Update error 0x80073712 is the error code, doesn't help much as it just means something is missing or corrupt. Here are the entries from CBS.log after attempting DISM installation of the update.

Exec: Processing complete. Session: 30939028_1294363939, Package: Package_for_ServicingStack_2350~31bf3856ad364e35~amd64~~17763.2350.1.0, Identifier: KB5008287 [HRESULT = 0x80073712 - ERROR_SXS_COMPONENT_STORE_CORRUPT]
Exec: Servicing Stack Upgrade operation will complete after TiWorker is recycled.
Failed to perform operation. [HRESULT = 0x80073712 - ERROR_SXS_COMPONENT_STORE_CORRUPT]
Session: 30939028_1294363939 finalized. Reboot required: no [HRESULT = 0x80073712 - ERROR_SXS_COMPONENT_STORE_CORRUPT]

 

[philc43]

Hello wdjenkins,

Looking at your CBS log file I can see that a manifest file is missing or corrupt.

2022-02-01 08:59:10, Info                  CSI    00000013 Warning: Unable to repair manifest for component ([l:93 ml:140]'amd64_023da5fb41dc3982e550c8b366276df5_31bf3856ad364e35_10.0.17763.1697_none_09e444ccd45d648f') from backups directory with disposition (2). A backup file may not exist or may be corrupt. Falling back to WU.
2022-02-01 09:01:09, Warning               CBS    Current tick count: 800 lower than last tick count: 1050. [HRESULT = 0x8007000d - ERROR_INVALID_DATA]
2022-02-01 09:01:09, Error                 CSI    00000014@2022/2/1:17:01:09.101 (F) Attempting to mark store corrupt with category [l:15 ml:16]'CorruptManifest'[gle=0x80004005]

The fix below will replace that manifest file for you.

Warning: this fix is specific to the user in this thread. No one else should follow these instructions as it may cause more harm than good. If you are after assistance, please start a thread of your own.

1. Download SFCFix.exe (by niemiro) and save this to your Desktop.
2. Download the file below, SFCFix.zip, and save this to your Desktop. Ensure that this file is named SFCFix.zip - do not rename it.
3. Save any open documents and close all open windows.
4. On your Desktop, you should see two files: SFCFix.exe and SFCFix.zip.
5. Drag the file SFCFix.zip onto the file SFCFix.exe and release it.
6. SFCFix will now process the script.
7. Upon completion, a file should be created on your Desktop: SFCFix.txt.
8. Please attach this file into your next post for me to check.

When this is done please run SFC /scannow which will ensure all the files are correct and then try to update again. If you get any errors please post the latest set of CBS.logs.

 

[wdjenkins]

Thank you so much, that worked! I ran the fix and did a SFC /scannow, but the update still failed. So I went for one more DISM /ONLINE /CLEANUP-IMAGE /RESTOREHEALTH and that was the trick. The update ran with no issues after that last DISM.

 

[philc43]

Good to know and yes that would have been the next step I would have recommended. Thanks for letting us know!