Siroj

Hi guys. We are having discussions on what path we should follow regarding the in-place upgrades of the laptops from the personnel that is working from home. All those laptops are connected through direct access and we have upgraded most of the laptops to build v1709 using a regular TS. Major difference was that at the time the personnel were at the office most of the time so we could pre-cache all data to the laptop when they were there. Now that almost all of our 3000 personnel is working mostly from home, the "pre-cache at the office" is not a valid option anymore. Alternatively we could download all data for the in-place upgrade through direct access but that's not what we want. We are looking at how to cope with this now. Ideas are making use of a DP in CMG and get the required data from there. Or perhaps a DP in our DMZ could be an option. What i am looking for is how you are managing this, what path have you chosen?

Our Setup: - SCCM 1806 on Server 2016 - Clients are running Windows 10 1709 (about 25000 clients) - Clients running corresponding SCCM client v5.00.8692.1008 We are trying to utilize Peer Caching but are unsuccessful in getting it to work. After a number of tries to connect to a peer, it uses the DP as last resort and successfully downloads the data (but not from a peer cache source). We verified that firewall ports (8003, 8004) are opened at the source peer. We re-entered the NAA account info and successfully verified. For testing purposes we also granted full access rights to CCMCACHE folder to the NAA account In SMSTS.log we see that the client has located a peer, tried to connect but failed. Some possible relevant LOG entries: Does anyone have any idea where to look next?

Take a look at: http://ccmexec.com/2016/08/disabling-windows-defender-whats-new-in-windows-10-1607/ That should help you out!

I know that it's possible to automatically set WinPE resolution by setting the X and YResolution variable when using MDT integration. Using an unsupported (low) value results in automatic detection. We want to incorporate the same solution, but without MDT integration. I can't find a solution to this, so my question is if this is even possible? I won't mind adding some custom variables or making other customization's.

We would like to create the following situation if possible: A new Windows 10 build/version is made available and is prepared by our technicians through SCCM. An upgrade TS is created and deployed to specific devices (laptops). When at work, an employee receives a notification through Software Center that an Upgrade Operating System (TS) is available. The employee selects the TS in Software Center. In the background all required components are downloaded/prestaged to disk and when ready the employee is presented with a message that the Upgrade is ready to be performed. At this stage it should be possible to go home or anywhere else and kickoff the upgrade, not having to be connected to the office network for the upgrade to complete. Is above scenario possible with current technology?

We are going to use SCCM to deploy Windows 10 v1607 to about 5000 devices. A script is written to detect the monitors resolution and copy a matching wallpaper to the device. What we want then is (preferably using GPO) to set this wallpaper as default to our users on first login and allow the user to change the wallpaper. We know we can set a fixed wallpaper using GPO, but that is not what we want/need in this case. If i use GPO Preferences to create the items Wallpaper and WallpaperStyle under [HKEY_CURRENT_USER\Control Panel\Desktop] and set it to only apply once, it gets overwritten with the default values while creating the profile after first login. So my question is, how can i set this wallpaper as default to our users on first login and allow the user to change the wallpaper?

We have a Windows 10 task sequence that uses a Apply Driver Package step. Drivers get installed correctly, as it seems when reviewing smsts.log. The correct package is applied. After the TS is complete, some devices have no driver installed. Now the weird stuff comes to play... When we wait a few seconds/minutes these missing drivers get installed automatically. The machine has no internet connection so the drivers have to be available to the OS. Why are these drivers installed after the TS is complete and not during the Driver installation step during the TS? And more important, how can we overcome this behavior?

I have created a new log and attached it in full to this post. Its a simple capture with 3 packages installed, nothing fancy. Notice the 80072ee7 errors around Prepare OS phase. smsts.log

The errors occur right after the sysprep phase and another time at the end of the TS. The network adapter used in the virtual machine is the generic E1000 and for what i was able to check, during WinPE and FullOS the adapter was having an IP address and was able to ping and resolve the FQDN. THe only thing i have to double check is if it looses it's netwerk connection during the prepare OS phase, the phase when the errors are written in the logs. I don't know if i am able to check the connection at this point and if i am fast enought to do so. Any tips on that? Is there any info on what's happening during the time the errors are logged, mainly things that could explain the loss of connection to the server/network? As it's a virtual environment in our ESX farm, there is no issues like waiting for Spanning Tree for example.

I am building a Windows 10 10586 reference image. The image builds without any noticeable faults and the captured .WIM file is available after the process is completed. When examining the smsts.log however, multiple 0x80072ee7 errors are present. All these errors occur after sysprep has been executed. I know this error is DNS/network related most of the time, but i have checked connectivity during all stages when possible and the connection is available, the FQDN is resolved and able to ping. All previous steps like installing software, installing configmgr client etc are processed without errors + the TS finishes without further errors. Any help is appreciated. Sample of error, smsts.log is attached with relevant section of log:

Issue is resolved. I have changed a couple of things so what fixed the issue is a bit of a guess but what i noticed was that Windows was being installed on drive X and partitioning was incorrect. For now i let Windows 10 install on a single 100% primary partition and i make use of the OSDPreserveDriveLetter variable.

I have a basic task sequence to capture a virtual (VMware ESX and/or Workstation) Windows 10 10586 reference machine with no applications or customization for testing purposes. The process completes successfully and .WIM file is available. After capturing the machine i reset the virtual machine to verify if installed machine is ok, but when rebooting it wont boot and is only showing the Windows logo and looping circle. Any idea?

At this moment we are in the process of creating a new Windows 10 image and all related stuff (TS, apps, etc) in our lab/test environment. This lab env is running SCCM Current Branch (1511). Our production env is running 2012 R2 and we are preparing and planning to upgrade this production env to the same level as our lab env. Both env are separate hierarchies with different site codes, but are member of the same domain and have full network connection. Data source of both env is stored on a central SAN that is available to both env. What we want is to be able to copy/export/migrate all relevant stuff from the lab env to our production env when this is fully tested and approved. It would be very nice if we would be able to accomplish this export from the newer Current Branch version into our production environment that is running an older version, but realistically i guess this is not supported nor possible and don't want to spend to much time into this scenario. What i do want to know is what's the best method of copying/exporting/migrating from our lab env to our production env? I know the existence of the TS export function and the migration tool, but have not tested both. Testing both is no problem, but i would like to know your experiences with this kind of setup. Any help/tips is greatly appreciated!

Bug is reported on Connect: https://connect.microsoft.com/ConfigurationManagervnext/Feedback/Details/2134832

New upgrade files appeared today with added "Volume" and "Retail" description, unfortunately still the same error when downloading. Patchdownloader.log shows the new entry refers to the same file as before.