Hello Everyone!
I am trying something in my test environment.
Following is the problem,
1. Windows 2016 Standard VM, running on vSphere
2. vSphere snapshots of this VM are replicated to S3, at regular intervals (approx. every 2 days)
3. We regularly do recoveries on AWS to test our backups
4. One thing i noticed is that most of the time when there is a Windows Update on this VM, it goes into Recovery Mode when recovered on AWS (Screenshot attached).
5. All the activities of conversion, driver needs, is taken care by our internal scripts and we see this problem mostly only when the source VM has System Updates.
6. I have confirmed (to the best i can) that the when the live VM snapshot was triggered, the System Updates were either downloaded, installed or the VM was rebooting after the updates were applied (as per the timestamps in Windows and vSphere logs.
1. For trying to debug these problems on AWS, i created a temp Windows instance and attached the failing OS disk to it.
2. When i try to run any DISM commands (to check health, get drivers etc.) for the failing OS disk, i get Error: 87 (screenshot attached)
3. When i try to run a sfc scan offline, i get the 'could not perform operation' message. (screenshot attached)
4. Chkdsk returned without any errors.
5. BCD enum also looks ok. Just for the sake of it, i also tried 'Bcdboot e:\Windows /s / D: /f BIOS' which succeeded. But when i tried to boot the instance with the failed OS, it was the same Recovery mode screen.
We are moving to a software which will manage the System updates automatically (so i cant control when or how they are pushed), and i also dont want that to hurt our backups.
Racking my head on this for a while, so any suggestions, approaches will be very highly appreciated.
Thanks!
I am trying something in my test environment.
Following is the problem,
1. Windows 2016 Standard VM, running on vSphere
2. vSphere snapshots of this VM are replicated to S3, at regular intervals (approx. every 2 days)
3. We regularly do recoveries on AWS to test our backups
4. One thing i noticed is that most of the time when there is a Windows Update on this VM, it goes into Recovery Mode when recovered on AWS (Screenshot attached).
5. All the activities of conversion, driver needs, is taken care by our internal scripts and we see this problem mostly only when the source VM has System Updates.
6. I have confirmed (to the best i can) that the when the live VM snapshot was triggered, the System Updates were either downloaded, installed or the VM was rebooting after the updates were applied (as per the timestamps in Windows and vSphere logs.
1. For trying to debug these problems on AWS, i created a temp Windows instance and attached the failing OS disk to it.
2. When i try to run any DISM commands (to check health, get drivers etc.) for the failing OS disk, i get Error: 87 (screenshot attached)
3. When i try to run a sfc scan offline, i get the 'could not perform operation' message. (screenshot attached)
4. Chkdsk returned without any errors.
5. BCD enum also looks ok. Just for the sake of it, i also tried 'Bcdboot e:\Windows /s / D: /f BIOS' which succeeded. But when i tried to boot the instance with the failed OS, it was the same Recovery mode screen.
We are moving to a software which will manage the System updates automatically (so i cant control when or how they are pushed), and i also dont want that to hurt our backups.
Racking my head on this for a while, so any suggestions, approaches will be very highly appreciated.
Thanks!
