anyweb

the firefox app is just a test old version here are some of the properties...

it seems to be working fine for me, here's my task sequence

ok i've fixed my certificates and i'm now seeing this on a very simple build and capture task sequence with just one application, firefox, is this basically the same error you are seeing ? or not ? i'm going to stick a pause just after this step to do some further troubleshooting.

i created the task sequence just now but it's not working as i've expired certificates in the lab, i have to fix that first, so please be patient.

it's soooooooooooooo long since i've done a build and capture, let me test it in my lab and post back here, forget about the lmhosts for now. i'll post back later

Introduction Microsoft pushed out some new updates for Windows 365 Cloud PC’s recently and one that we thought was interesting was this one, the ability to recover a Cloud PC that was de-provisioned due to license expiration. The de-provisioning of a Cloud PC normally means that it is unusable and unrecoverable, it’s a destructive action. However, with this new feature and as long as you have previously enabled Point-in-Time-restore explained here then you should be covered by this new ability (for Windows 365 Cloud PC Enterprise only). In this blog post myself and my good friend Paul look at the new feature and test it out. We’ll be collaborating on more Windows 365 and Azure Virtual Desktop content during the year so stay tuned on the SCCMentor and windowsnoob websites! The prerequisites Before we test it, let’s look at what you need to have in place. To recover a previously de-provisioned Cloud PC, ensure that the expired Windows 365 licenses are renewed. The replacement licenses must be of the same SKU, so if the previous Cloud PC’s were Cloud PC Enterprise 2vCPU/8GB/128GB then the replacement license assigned must be for Cloud PC Enterprise 2vCPU/8GB/128GB. To recover a previously de-provisioned Cloud PC, ensure that the provisioning policy of the de-provisioned Cloud PC still exists. To recover a previously de-provisioned Cloud PC, ensure that assignments with the right users are assigned to the provisioning policy. To recover a previously de-provisioned Cloud PC, ensure the end users still exists, and are still assigned to the same Entra ID groups as before or the ones included in the policy assignments. Testing the new feature Ok now that’s all clear, let’s test it out. Before starting, we verified that our test Cloud PC had a point-in-time restore policy (User settings) applied. We also made sure to place a file on the Cloud PC outside of Onedrive known folders. And before continuing, we verified that there was at least one backup for this Cloud PC, and there were several. Note: If you are also testing this feature and you are doing it on a newly provisioned Cloud PC, you can click on Create restore point to create a manual restore point. As simulating a license expiry and then re-purchasing the same license is easier said than done we’ll simulate it by deleting a user that was assigned the following Cloud PC SKU: Cloud PC Enterprise 2vCPU/8GB/128GB. Below you can see that user (testuser100@windowsnoob.com) and their respective Cloud PC. Next, we deleted our licensed user. and were notified of that… Shortly after, the status of the Cloud PC that that user was using, changed from Provisioned to In-grace period. Clicking that In-grace period status and selecting Deprovision now, speeds up the testing process. Which kick starts de-provisioning Some time later, the Cloud PC was no longer in the Intune console and was no longer available to the (now deleted) end user to use. At this point, the de-provisioning process was completed. Recovering de-provisioned Cloud PCs In the real world, to see this new feature you’d renew an expired license and make sure that the user/group/policy and so on didn’t change during that time period. But we are simulating that, so to simulate it we’ll restore the user we deleted earlier, again, nothing else has changed other than that users Cloud PC’s went from in-grace period to de-provisioned. So let’s restore the deleted user. In Azure (Entra Id) select users, click Deleted users and locate your deleted user. Select the user and click Restore users. You’ll be notified of the restore. And as the user was still in the same groups, they’ll pick up the Cloud PC assignments targeting them and a new Cloud PC matching the SKU in question will be provisioned. After the Cloud PC is provisioned, you’ll see it appear with a new computer name. We need to do some final actions. Select the newly provisioned Cloud PC and click the Restore option. This will show a retention restore point. Select it and click the Select option to restore that de-provisioned Cloud PC. You’ll get a popup warning you about the restore. Click Restore and the restore will be initiated, you’ll see the status change to Restore pending. And during the restore the status will then change to Restore active. and after a while refreshing that device will reveal a ‘Not found‘ statement (sorry no screenshot, we’ll add it shortly). However, if you navigate back to Windows 365 devices, you’ll see that the old name (old Cloud PC) is back ! and logging on to that Cloud PC using the same user will indeed reveal that the content is as expected. Job done! Read more To learn more about this new feature see Microsoft’s own page here. To learn how to use Point-in-time-restore read here.

Introduction Look out for a new series of blogs from myself and my MVP buddy, Paul Winstanley as we explore the world of Azure Virtual Desktop (AVD). In the meantime, we have collaborated on a kick the tyres exercise to experience the AVD solution via the Microsoft quickstart. This blog post details that experience with a simple step-by-step guide. Microsoft provides the quickstart as a great way to test out an AVD environment with minimal fuss. It will create a small, low cost environment in approximately 20 minutes. Prerequisites To create the environment you will need to meet the following prerequisites: An active Azure subscription An Entra account with the following roles assigned: Contributor User access administrator Some test Entra accounts, for authentication into the AVD environment Available quota for your subscription for the Standard_D4ds_v4 virtual machine. If you have insufficient quote then you can increase – https://learn.microsoft.com/en-us/azure/quotas/per-vm-quota-requests Access to the required FQDNs and endpoints in Azure – https://learn.microsoft.com/en-us/azure/virtual-desktop/required-fqdn-endpoint?tabs=azure What gets deployed during the Quickstart A resource group called rg-avd-quickstart-<GUID> A virtual network and subnet. The IPv4 address space will be 192.168.0.0/24 and will use Azure provided DNS servers A network security group that is associated with the subnet of the virtual network and uses only the default rules. No inbound rules are required for Azure Virtual Desktop A host pool with single-sign-on (SSO) enabled An Entra joined session host running Windows 11 Enterprise multi-session with Microsoft 365 apps preinstalled with English (US). The virtual machine is a Standard_D4ds_v4 size virtual machine (4 vCPUs, 16 GiB memory) configured with a standard SSD disk An application group that publishes the desktop from the session host A workspace For a complete list of the resource published by the quickstart, take a look the Microsoft documentation https://learn.microsoft.com/en-us/azure/virtual-desktop/quickstart?tabs=windows#deployed-resources. Getting started The first step is to register the Microsoft.DesktopVirtualization resource provider on your Azure subscription In the Azure portal, go to your Subscriptions and select that one you will use for this exercise. Select Resource providers from the menu and look for Microsoft.DesktopVirtualization. If this is listed as NotRegistered in your subscription, then click the three dots and select Register as shown below. If it is already listed as Registered then there is nothing else you need to do. Before continuing, ensure that Microsoft.DesktopVirtualization is listed as registered. Deploying the quickstart solution You can find the quickstart solution in the Azure portal by searching for Azure Virtual Desktop and then selecting Quickstart from the menu. Click the Create button to begin the process. You will need to enter specific information on the Basics tab. Subscription: Ensure you choose the subscription where you have registered the resource provider previously. Location: Select the Azure region where you want the AVD resources deployed to. Local administrator account: Username: Enter the name for a local administrator Password: Enter and confirm your password for the local admin User assignment: Select a maximum of two users who will have access to the AVD devices (see screenshot below) Once all the details have been entered click Review + create to continue through the quickstart wizard. Ensure you receive a confirmation that Validation passed and click Create. The deployment of the resources will be in progress. If all goes well, Your deployment is complete will be returned. Connecting to your AVD desktops You can connect to your AVD desktops via the Windows app, which is available from the Microsoft Store. When installed, and one of your allocated users is authenticated to the app, you should see a SessionDesktop item (in addition to any Cloud PC’s or other AVD’s available to that user). This is your AVD desktop. Click Connect to access. Once done, you will be prompted to enter your password, or MFA details and then asked to Allow remote desktop connection. This message can be suppressed using a list of trusted devices (see https://learn.microsoft.com/en-us/azure/virtual-desktop/configure-single-sign-on for more information). Just to note, you can also alter the display settings for your AVD session by clicking the ellipsis (three dots) and adjusting. Eventually, you will hit the Windows 11 desktop. Since this is multi session, your other allocated user can log in and use the AVD desktop at the same time. Note, the screenshot below, the same hostname is in use. Removing the quickstart services Once you have used the quickstart, you may wish to clean up the resources, ready for your deeper dive into AVD. To do this, locate your AVD virtual machine under Virtual machines in the Azure portal and click on the device. Click Stop to deallocate the device. You will be prompted to confirm the deallocation. Make sure you have ony selected the AVD virtual machine that you want to stop. Click Yes. Wait until the device reports back as Stopped (deallocated) under its Status. Now we need to remove the resources which were created, as they encur costs when running. You can do this 2 ways, in Azure, search for Resource groups and find the one created by AVD. As mentioned, this will have a naming convention of rg-avd-quickstart-<GUID>, or simply click the AVD virtual machine and then select the resource group it’s in as shown here. Click into the resource group, select all the resources under Overview and then click Delete. You will need to confirm their removal by entering delete in the field below and also you can select Apply force delete for selected Virtual machines and Virtual machine scales sets, if you so wish. Once done, click the red Delete button. You will then need to confirm the deletion of the resources. Click Delete one more time. The resources will no longer be listed when they are removed. Next, click Delete resource group to remove the quickstart resource group. Enter the name of the resource group to confirm its deletion and click Delete. Once again, confirm by clicking Delete. I hope this guide gives you a good oversight into how quickly you can test out the Azure Virtual Desktop using Microsoft’s quickstart option. You can get a VM up and running in no time and experiment with it. As mentioned, we will be taking a deeper dive look at AVD very soon, so watch this space and keep an eye on the SCCMentor and windowsnoob websites.

great findings and glad you got it sorted are you sure with the changes made that it's going to work security wise ?

send me a pm

I would try yes, but can i guarantee no issue, no. if you do get a problem post here and i'll do my best to help

Using CMTrace error lookup, I can see that the error message is actually an RPC server is unavailable I asked Google (ai) what this could be, The "RPC server is unavailable" error (0x800706BA) in SCCM (System Center Configuration Manager) typically indicates a problem with communication between the SCCM server and a client machine. This usually stems from network connectivity issues, firewall restrictions, or incorrect service configurations. Here's a breakdown of potential causes and solutions: 1. Network Connectivity Issues: Verify Network Connectivity: Ensure that the SCCM server and the client machine can communicate over the network. Try accessing the client machine from the server using its hostname or IP address in Windows Explorer (e.g., \\clientname\c$ or \\192.168.1.10\c$). Check DNS Settings: Incorrect DNS settings can cause name resolution issues, preventing the server from finding the client. Verify that the DNS settings are correct on both the server and client. Time Synchronization: Ensure the time and time zone settings are synchronized between the client and server. Inaccurate time can also cause communication problems. 2. Firewall Restrictions: Firewall Configuration: Firewalls, both on the client and potentially on the network infrastructure, might be blocking the necessary ports for RPC communication. Specifically, verify that ports 135 (TCP and UDP) and the dynamic RPC port range are open. Outbound Traffic Inspection: Some firewalls with outbound traffic inspection capabilities might be blocking RPC traffic. Check if any such rules are in place. 3. Service Issues: RPC Services: The Remote Procedure Call (RPC) service and other related services, like the "DCOM Server Process Launcher" and "Remote Registry" services, must be running on both the client and server. Service Configuration: Ensure that these services are set to start automatically (Automatic) after a system restart. 4. Account Permissions: User Account Permissions: Verify that the account used for the SCCM connection has the necessary permissions to access the client machine and its resources. DCOM Permissions: Check DCOM (Distributed Component Object Model) permissions on the client machine, as they might be preventing the connection. 5. Troubleshooting Steps: Restart RPC Services: Restart the RPC service on both the client and server to reset its state. Reboot the Client: After making configuration changes, reboot the client machine to ensure the changes are applied. Use wbemtest.exe: Test WMI connectivity to the client machine using the wbemtest.exe tool. This can help isolate whether the issue is WMI-related or broader RPC communication problem. Deploy Recast Agents: If using Recast Software's Right Click Tools, deploying their agents can help resolve RPC communication issues. In Summary: The "RPC server is unavailable" error in SCCM is often a network or configuration issue. By systematically checking network connectivity, firewall settings, service states, and account permissions, you can identify and resolve the problem.

but you are already running SCCM 2403, right ? so... just go ahead and upgrade

You can upgrade directly to 2503 it's cumulative. So it should contain all the fixes and updates from the previous version. do check out the official known issues here: https://learn.microsoft.com/en-us/intune/configmgr/core/plan-design/changes/whats-new-in-version-2503 and my take here: https://www.niallbrady.com/2025/04/04/configuration-manager-2503-is-out-350-bugs-fixed/

I would escalate this to your Dell Technical Account Manager, this is not an SCCM problem,unless of course your quote here: was when you imaged the Dell using your own task sequence, is that the case ?

check the schedule is set to 'always rerun program' like below

ok in my 2503 lab it's working fine, just checked

i'll double check my 2503 env and let you know if it also has the same problem (or not)

well my CMG from the post above fixed itself over the period of 2 weeks after posting this blog post. I.e. it was a back-end fix from Microsoft and nothing that I did could change that. If you've made sure you have the latest updates/kb's applied to your 2503 environment and it's still showing this problem, try the fix mentioned above, if it helps great, if not hopefully microsoft will again fix it in the back end.

please see https://www.niallbrady.com/2020/09/15/fixing-an-evaluation-version-of-ssrs-with-http-error-503-the-service-is-unavailable/

Introduction Microsoft recently released a new feature for Windows 365 Cloud PC’s namely the ability to move one or more Cloud PCs to another region. This involves editing a previously created provisioning policy and then deciding whether you want to move all Cloud PC’s or only a subset targeted by that policy. All data/settings etc on the device (except for stored snapshots) will be retained with the move, which is great from an end user perspective. Microsoft themselves recommends testing this on a few Cloud PCs to review how the end to end process works for your users’ Cloud PCs. Why change location ? But first of all, why would we want to change the location of a Cloud PC ? Changing the location of a Windows 365 Cloud PC can be beneficial for several reasons: Improved Performance: Moving the Cloud PC to a location closer to the user can reduce latency and improve overall performance, making applications and services more responsive. Compliance and Data Sovereignty: Different regions have varying regulations regarding data storage and processing. Moving the Cloud PC to a location that complies with local laws can ensure adherence to data sovereignty requirements. Cost Optimization: Some regions may offer lower costs for cloud services. By relocating the Cloud PC to a more cost-effective region, businesses can reduce their operational expenses. Disaster Recovery and Redundancy: Relocating the Cloud PC to a different region can enhance disaster recovery plans by ensuring redundancy and availability in case of regional outages or disasters. User Experience: If a business has employees in different geographical locations, moving the Cloud PC closer to the majority of users can enhance their experience by providing faster access to resources. Scalability: Certain regions may offer better scalability options or more advanced infrastructure, allowing businesses to expand their operations more efficiently. Security: Some regions may have more robust security measures or offer specific compliance certifications that are important for certain industries. So let’s dig in and see how this works in reality. Moving one or more Cloud PC’s In the Intune console, locate the Windows 365 node and select Provisioning policies. Select the target provisioning policy, and click it to edit it. the policy details are revealed. Take note of the current location configured within the policy, this is optional but useful to know if you want to revert to that location later on. Review location before change At this point, pick a test Cloud PC to allow you to review that the move process goes smoothly. On that Cloud PC, verify the current location using a site such as https://www.whatismyipaddress.com or https://mylocation.org/. Below you can see the approximate location of our test Cloud PC before any move is attempted. As we can see it’s listed in Iowa, in the United States. 1. Inform the users Now that we know which provisioning policy we’ll be editing, and what location our target Cloud PC’s reside in, we need to pick one or more users of those Cloud PC’s and inform them of what is about to happen (and why). You need to explain to the users that they must save any open documents, close their apps and to sign out of their Cloud PC for the weekend. It’s a good idea to do the move over the weekend as obviously some Cloud PC’s will have more data/apps/settings to move than others. 2. Edit the policy In the General section of the provisioning policy, click Edit and modify the Geography. I chose European Union as that’s where I want to move my target Cloud PC’s. You can fine tune the region and select an actual region such as Sweden, but it’s recommended to choose the Automatic (Recommended) option to avoid provisioning failure. For a list of supported regions see here. As this is an actual move and not a provisioning, I chose Sweden. after editing the policy, click Next followed by Update. 3. Pick one or more target Cloud PC’s Now that you’ve changed the Geography and Region of your provisioning policy, click Apply this configuration to select one or more target Cloud PC’s. To select one or more devices, click the third option which is Region or Azure network connection for select devices and click Apply. That will bring up a list of devices to select. Select one or more targets and place a check in the Cloud PC’s will be disconnected and shutdown while the configuration is updating. Any unsaved work may be lost message. Note that the disconnect only affects selected devices, not all targeted by this provisioning policy. 4. Review the Cloud PC’s status The status of the selected Cloud PC’s will change to Moving region or network in the All Cloud PC’s section of the Windows 365 node. Individual Cloud PC’s will also show new status in Intune devices via the Overview and Device action status areas of each device. Also to note, the above statuses will have changed even before the move of the actual target takes place, so the Cloud PC may still be online for a brief period before it receives the instructions to shutdown. You can also monitor the status using the Cloud PC actions report. which reveals some more data for individual Cloud PC’s plus, if you selected multiple Cloud PC’s, click on Bulk batches. After some time, the status should update on the device to Completed and the Cloud PC actions report will be updated with the new status. And on the target PC itself, you can verify it’s location using the previously mentioned sites to confirm the new location. Job done! Final thoughts Keep in mind that after you’ve completed your Cloud PC move operations, that any new Cloud PC’s targeted by that provisioning policy will also be provisioned in the new region. This new ability is great, and I’ve tested it successfully in multiple tenants. I do however feel that with just a little bit more work it could be even better. What I’d like to see is native ability within Intune to send customizable emails/alerts/notification for any Cloud PC’s targeted by the move operation, both before and after the event to alert the end users about what is happening and when, and more importantly to let them know that operations are complete. Great job Microsoft!

Introduction In a previous blog post I started troubleshooting why my Windows app showed a blank white screen instead of the usual User Interface. I did a bunch of troubleshooting but it didn’t lead anywhere or so I thought. No matter what I did, I could not use the Windows app any more. So I kept digging. By checking the Details section of Task Manager, I could see where the Windows365.exe executable file was launched from. C:\Program Files\WindowsApps\MicrosoftCorporationII.Windows365_2.0.366.0_arm64__8wekyb3d8bbwe I browsed the files and folders in there and found an interesting to me json file called system_settings.json in the Resources folder, it referenced the Microsoft remote desktop client agent (msrdc) and a quick search on the internet revealed that that version was not the latest. So I downloaded the latest version and installed it. I then re-attempted to start the Windows app, but to no avail, it was still blank. The MSRDC agent I’ve been using the Windows app since it’s first iteration, so it was easy to forget about msrdc. Launching that Microsoft Remote Desktop Client agent allowed me to see and use my Cloud PC’s so that was at least a workaround for accessing them until the problem was resolved. Back to the problem though. Looking through event viewer I could see lots of errors in AppXDeployment-Server, maybe they had something to do with my problem. That error was: Deployment Register operation with target volume 😄 on Package MicrosoftCorporationII.Windows365_2.0.365.0_arm64__8wekyb3d8bbwe from: (AppXManifest.xml) failed with error 0x80073D02. See http://go.microsoft.com/fwlink/?LinkId=235160 for help diagnosing app deployment issues. I followed the link and it revealed the following for that error code. Webview2 revisited This ‘currently in use’ error got me thinking again about the Outlook New Webview2 problem, which I never managed to solve in part 1. But first, what is Webview2 ? see here. WebView2 is a way for app developers to embed web content (such as HTML, JavaScript, and CSS) in Windows applications. By including the WebView2 control in an app, a developer can write code for a website or web app, and then reuse that web code in their Windows application, saving time and effort. See Introduction to Microsoft Edge WebView2. So what was the Outlook New problem ? Namely every time I started Outlook New, it would prompt me to install Webview2. As you can see in the previous part I tried to fix that and thought that I had, but in reality, the problem was still there and Outlook New always complained about the Webview2 requirement. Could the Webview2 installer be stuck in a loop somehow and that could be affecting the Windows app from updating/deploying properly ? When I tried to solve it in part 1, I launched Outlook New from an administrative command prompt, or chose “run as administrator” after elevating myself. However the solution we use for elevation only elevates processes such as cmd.exe or what you click on. It seems that the Webview2 installer, ignores the permissions that Outlook New is launched from. Eureka This became clear when I forgot my phone at home, and thus was not able to enter the MFA code to elevate a process in my continued troubleshooting., instead, I used ControlUp to elevate my session. I then logged off, and logged on to Windows again, and tested my elevation, I was elevated. The difference here is my Windows session was elevated, not just a process. I then launched Outlook New as administrator, got the Webview2 popup, clicked OK and …. some minutes later, Outlook opened, for the first time in days! Clearly, Webview2 was now installed. I then launched the Windows app using ‘run as administrator’ and it too, launched, successfully ! Wonderful ! So at this point I’m fairly confident that Webview2 was the problem here, or if both the Windows app and Outlook New for some reason needed the Windows session elevated in order to complete pending Webview2 update actions. I’m also very happy that my Windows app is once again working on my Qualcomm ARM64 laptop ! If you are wondering whether Webview2 is even used for the Windows App, open task manager, search for View and launch the Windows app. Look what appears. I’ve placed red rectangles around three things, Outlook, Windows 365 and MS Teams. Why the other apps ? during my the last week or so I also had serious lagging issues in Teams, and it too, uses Webview2 as you can clearly see in the image above. Fixing the Webview2 installer in Outlook (New) via an elevated Windows session solved all my problems in the following apps: Outlook (New) Teams Windows App Update I’ve since had the exact same problem occur another 3 times and I’m getting much better at fixing it, especially after reading this. In a nutshell elevate your Windows session (as described above), then delete the following registry key: HKLM\SOFTWARE\WOW6432Node\Microsoft\EdgeUpdate\Clients\{F3017226-FE2A-4295-8BDF-00C3A9A7E4C5} Once done, run the Edge troubleshooter here. This will successfully download and reinstall Webview2 and life will be bearable again, for a while. I’m not sure what app or process breaking my Webview2 but I can tell you this much, it’s very very annoying.

Introduction I use my Windows app daily from a variety of devices to access my Cloud PC’s. Today like any day I clicked the Windows app icon but this time, nothing happened, the familiar Windows app did not appear. So I clicked it again and this time it did appear but was blank. Note: This problem occurred only on my ARM64 device, the Windows app worked just fine on other X64 devices. As you can see in the screenshot above, it’s completely blank, empty of any content, so it’s impossible to use. Trying to fix it I thought the app somehow became corrupt, so I tried to reset & repair the app. Neither of these options helped. Next, I tried to end the processes for the Windows in task manager. And killed the Windows365.exe process tree… but these didn’t help either. The app simply refused to display it’s content. While troubleshooting I noticed a Windows cumulative update was ready to install, you can see it below with the blank Windows app in-front. so I let Windows update do it’s thing and manually restarted the computer. It didn’t help. What about the logs ? Next I looked at the logs connected to the Windows app in %temp%\DiagOutputDir\Windows365\Logs as documented here. The logs (4 in total) did not give me anything to go on other than to highlight that I did have a problem. The most interesting log snippet was this” [2025-03-21 12:53:49.547] [async_logger] [error] [MSAL:0003] ERROR ErrorInternalImpl:134 Created an error: 8xg43, StatusInternal::InteractionRequired, InternalEvent::None, Error Code 0, Context ‘Could not find an account. Both local account ID and legacy MacOS user ID are not present’ but I saw that was repeated even when my Windows app was working, so I chose to ignore it. So what else could I do at this point ? uninstall the app. I then reinstalled the app from the store. end result, same problem! At this point I reverted back to the logs, and found this section, it reminded me of an error I saw this week when launching Outlook (New). The Outlook (New) webview2 error is also shown on top of the log file below: I never bothered to install that Webview2 component as it requires local administrative permissions to do so and I’m a standard user (as I should be). So I elevated my session and tried clicking ok to the Outlook (New) message. That didn’t help, it just kept looping through a prompt asking me to install the new Webview2. A google search led me to this post, and I downloaded the Webview2 client (for ARM64) from here. After deleting the reg key and installing the Webview2 update & restarting the computer I still had the issue with Outlook New prompting to install Webview2 and my Windows app not working. Next up I checked the version of Windows app I was running and it was the latest, 2.0.365.0(ok there’s a insider preview version that’s newer but… I don’t think that’ll help). At this point I’m out of ideas and will inform the Microsoft Group in charge of Windows 365 to see what they have to say about it, I’ll revert in part 2 once this is hopefully solved cheers niall

can you share your smspxe.log file with me and i'll take a look, it may reveal the source of the problem,

it could simply be corrupt task sequence policy, can you do a test and deploy the exact same TS to either the same collection or another collection that this device is in, and retry I tested pxe boot just now in my 2503 lab and all is good, even in a fully PKI environment

the last error in your provided log is "There are no task sequences available to this computer.. Please ensure you have at least one task sequence deployed to this computer. Unspecified error (Error: 80004005; Source: Windows" so.. make sure that whatever you are pxe booting actually is in a collection that is targeted by a task sequence