Hostname Changer AD Issues
-
@lambo I am not exactly sure what size and file types are allowed in the forums. You can give it a try or upload to a file share/cloud of your choice and post a link here.
-
This post is deleted! -
Here is a link you can try, Let me know if you are able to access the drive.
https://drive.google.com/drive/folders/1aOwYRK71pVZL68uIrzPzT1ylOmquVzeP?usp=sharing
Fog logs are listed, Testing Notes.txt are the overall notes showing what tasks have been taken for each fog log.
Let me know if you need anything else!
Thanks!
Lambo
-
@lambo Well done. Looking through your notes and logs I find it very awkward that we have different results for tests that should end up the same! The fog-client simply calls official MS API to do rename, AD join/leave and so on. While it’s still possible that it’s something we do wrong in the code I find it very strange that it would work one time and fail the next few tries.
What I found in the fog-client logs is a few unknown return codes:
FOG Test 1.log: 2/25/2021 9:10:45 AM HostnameChanger Unknown Return Code: 1791 FOG Test 5.log: 2/26/2021 10:19:55 AM HostnameChanger Unknown Return Code: 1791 FOG Test 7.log: 2/26/2021 1:34:39 PM HostnameChanger Unknown Return Code: 1791 FOG Test 8.log: 2/26/2021 2:26:16 PM HostnameChanger Unknown Return Code: 1332 FOG Test 8.log: 2/26/2021 2:28:48 PM HostnameChanger Unknown Return Code: 1332 FOG Test 9.log: 2/26/2021 3:49:57 PM HostnameChanger Unknown Return Code: 1791- 1791: A remote procedure call is already in progress for this thread. (reference)
- 1332: No mapping between account names and security IDs was done. (reference)
Possibly you have some other software interfering with the fog-client? Do you have other tools installed that would run straight after the client comes up after imaging?
Did you find anything obvious in the event logs on the client or the AD server when it fails to create the machine account?
As a side note: You still have the 0.11.19 fog-client installed. Unfortunately the CA certificate shipped with that binary ran out last year and so the auto updater can’t pull up to the 0.12.0 version coming with FOG 1.5.9. It downloads the SmartInstaller but cannot verify the signature and therefore cannot install it. So at some point you’ll need to update your base image with the newer fog-client.
Though I don’t think 0.12.0 will make much of a change in this scenario you are facing here. There have not been updates to the HostnameChanger module since 0.11.19.
-
Sorry for the delay, we have a major project that I am currently involved with in my organization.
I agree! But there must be a cause whether its on our end or something that may have gotten corrupted on our server. This whole issue is quite strange and is very perplexing. I can’t seem to find any rhyme or reason.
I will have a look at these error codes and see if I can’t find anything that would be causing this issue.
The only thing that we have is a snapin for our software deployment software that is set to run after the PC is imaged and on the domain. I can certainly try to run some images without this snapin enabled and see if that makes a difference?
Other than that, i am not seeing anything that could cause issues like we are seeing. I’m not seeing too much in the event logs and I don’t have access to the AD server, but i had the team look at it and they said they haven’t seen anything unusual.
Let me take a look at the Fog client as well, even though you might not think it could be the issue, we can try to ensure that it isn’t causing some funky issue.
Do you suggest maybe downgrading the server version to 1.5.8 as well just to test and ensure we didn’t encounter some weird issue when upgrading, perhaps something got corrupted on our end or similar?
Thanks, sorry again for the delay!
Lambo
-
@lambo said in Hostname Changer AD Issues:
Do you suggest maybe downgrading the server version to 1.5.8 as well just to test and ensure we didn’t encounter some weird issue when upgrading, perhaps something got corrupted on our end or similar?
Don’t think a downgrade will be of help in this case. Also be aware that going back is not something we officially support within the FOG software. It’s hard enough to make updating work for as many people as possible without too much trouble. So it’s a manual process of taking an old DB backup and so on. As well there is no auto-downgrade to an older version of the fog-client.
-
Ok, that is understandable. Is there anything that can verify the installation?
I will work on the Fog client and seeing if upgrading changes anything, as well as imaging without the snapin enabled.
Do you have any other suggestions?
Thanks!
Lambo
-
@lambo Will be interesting to see if imaging without the snapin does make a difference!
Other than that I don’t have other ideas right now. Please give that a try and we’ll see from there.
-
Certainly!
My team an I are working on running without the snapin, as well as updating the Fog client just in case!
I will be in touch.
Have a great day!
Lambo
-
Unfortunately, removing the snapin from the process and upgrading the Fog Client had no change in the outcome. We ran images all day yesterday with the same outcome.
I think I may try deploying a new Fog server and test if it has the same outcome. I don’t have any other ideas on our current server. Do you have any other suggestions?
Thanks!
Lambo
-
@lambo While I do understand the urge to rule out certain things to find out what is causing this I don’t think that setting up a new FOG server will be of any help. This is an issue where the FOG server itself is not involved really. Sure the fog-client polls the AD information from the FOG server DB but after that it’s just the fog-client calling Windows API and talking to your Windows AD server(s). If we would see error codes that tell us credentials are wrong, then we could blame or look into the FOG server but from the evidence we have right now I can’t see the FOG server causing this.
Would you be able/allowed to setup your own AD server for testing?
-
That makes sense! I was thinking along the lines of a corruption or similar with this specific issue.
I can check with our higher ups to see if we could create a local AD server for testing purposes to see if we would be able to work with this server on narrowing down the issues that are presenting themselves. Let me get back to you on this point and let you know if this is possible.
Thanks!
Lambo
-
I just wanted to update you Sebastian.
This Wednesday / Thursday I will be working with the AD team to view logs during the imaging processes. They are most likely going to request that I have access to set up a temporary AD Server / get access to their Test AD servers.
I am hoping to have an update for you after testing this week.
Please let me know if you need anything else.
Do you also have a link to the Fog recommended image process so i can ensure it is not an issue with our image as well? Our Windows version is Win 10 Enterprise 20H2.
Thanks!
Lambo
-
@lambo Sounds good. Keeping my fingers crossed that you will find out more soon.
Don’t think we have an official document on the imaging process. Just a few things that I have in mind:
- Disable secure boot
- Set disk controller to AHCI mode instead of RAID mode
- Disable bitlocker (
manage-bde -off c:) - Un-join from domain
- Disable fast boot in Windows 10 and do a clean shutdown
@george1421 @Tom-Elliott Some essential point I am missing?
-
Hi Sebastian,
I hope all is well with you!
Sorry for the delay.
Ok so we were able to produce the AD issue with the AD team.
Unfortunately, we didn’t see much happening in terms of event logs.
There were three separate failures that occurred, but not necessarily related to joining the domain in my experience. I could certainly be mistake though.
-
Audit Failure - Kerberos Authentication Service - 4786
-
Audit Failure - Directory Service Access - 4662
-
Audit Failure - Credential Validation - 4776
None of these failures utilized our image account for Fog / Hostname Changer, so I don’t see them as being a smoking gun unfortunately.
The AD Team was planning a migration for our AD server to Windows Server 2019, so they will be completing this ahead of schedule for us. This migration will take place this week, so we are hopeful that it could assist with resolving the issues that we are seeing.
If that does not work, the AD team will allow us to utilize a test AD server for troubleshooting this issue. We should then be able to fix up and test any settings we think are important.
Do you also know if it is possible to ‘lock’ the domain controller to a specific domain controller in Fog? This would be helpful so we could lock Fog in to utilizing our local domain controller instead of possibly utilizing another in the network.
Again, sorry for the delay, and thank you so much. I really appreciate your assistance!
Thanks!
Lambo
-
-
@lambo said in Hostname Changer AD Issues:
Do you also know if it is possible to ‘lock’ the domain controller to a specific domain controller in Fog? This would be helpful so we could lock Fog in to utilizing our local domain controller instead of possibly utilizing another in the network.
Do you have serveral AD controllers and want to force the client to use a specific one? That’s not something I can answer. From what I know about Windows AD I would guess the answer is No. But that’s definitely beyond my skills.
My thinking was you would setup a single AD testing server using a complete new domain for testing, not part of the existing domain in any way. Not sure if that’s possibly in your environment.
-
Hi Sebastian,
No worries at all, i just figured i would ask, as there are multiple domain controllers in our organization and i was hoping there would be a way to point fog to utilize the local DC. No worries though at all!
After testing with the Server 2019 DC in production, it seems that some of the issues presented have been resolved in some regard.
Here is what we found:
If a computer AD account already exists, it seems we can reimage with no issues now. we tested around 15 images and they all worked successfully. We can also reimage a computer where the AD account has been ‘reset’ instead of deleted. These tests were successful as well.
However, if we delete the computer AD account or image a new PC and have the option to add the computer to the domain, the issue we were previously experiencing exposes itself. The computer will allow AD accounts to log in, but trying to deploy software shows a trust relationship issue as well as having the computer nowhere to be found in Active Directory.
What would you like to see done here? I am about out of ideas on our end unfortunately, other than standing up a test domain and seeing if the issue persists there and if so, trying to narrow down where the problem is. I think we should be able to set this up and we can continue deeper testing there.
Thanks!
Lambo
-
@lambo Just wondering if you are still looking into that issue or got it solved at some point?