Known issues and considerations

2607 LTSR

• On Windows Server 2022 target devices, some devices randomly becomes unresponsive on a dark coloured screen after reboot. The target device software then remains stuck in reconnect mode.
• When you run Provisioning servers on XenServer, Streaming Service sometimes causes high CPU usage or CPU spikes. This issue appears on some servers while others in the same environment remain unaffected.
• When you add Cisco AMP or Microsoft MDE to the vDisk image and boot many targets, Stream process CPU usage climbs and stays high. Boot times increase, sometimes to around 10 minutes per target.
• When you update Azure hosting configuration from the PVS console in a hybrid environment, The following error message appears:
  Invalid Connection Settings
  You then can't manage Azure VMs from the console.
• When you run XenApp 2402 CU3 on-premises, multiple target devices on the same host sometimes fail to boot. The affected servers remain in a unresponsive state.
• After a scheduled reboot, some targets remain powered off and require manual intervention to start. Affected machines power on but fail to register with the Delivery Controller and then enter a fault state.
• After you upgrade to 2507 CU1, non-persistent PVS VDAs from the same image intermittently report incorrect load index values in Studio and Web Studio. Affected servers show session count only, not full evaluator metrics. After restart, results vary by server.
• On Windows 11 24H2 targets provisioned with Citrix Provisioning, the first Quality Update after a Feature Update causes streaming-mode boot failure. Devices become unbootable after reboot. As a workaround, apply the Quality Update through reverse imaging on local disk.
• When you attach the write cache disk at SCSI slot 0:0, write cache falls back to the PVS server instead of local disk. The same disk works at slot 0:1. This issue affects Windows Server 2022 and Server 2016 targets.
• In some cases, Stream process on Provisioning servers consumes high CPU and pegs server resources. Many targets then fail to boot. SQL Server might also spike CPU due to related queries.
• During scheduled reboots in Citrix DaaS with on-premises VDAs, some machines intermittently remain in a shutting down state. The guest OS receives shutdown, but force shutdown does not complete, and no further hypervisor tasks appear.
• Target devices intermittently enter licensing grace period about every 48 hours. The following error message appears:
  No device license is currently available for this computer. A system shutdown will be initiated in XXX hours.
  To avoid data loss, save work and close applications. As a workaround, run Configuration Wizard to temporarily resolve the issue.
• On the PVS server, notifier and inventory processes intermittently might become temporarily unavailable and generate Event IDs 155 and 158. The notifier process fails while opening SQL Server using Windows Integrated Authentication. The inventory process fails with an access violation.
• On Windows Server 2019 targets running Provisioning 2507, target devices sometimes remain in reconnect mode for long periods. The issue occurs randomly, shows no clear vDisk change, and persists regardless of third-party security software.
• When one or more license server fields in the PVS database are empty, the server cannot validate entitlement and might report as unlicensed. This occurs because the license entitlement client fails to build.
• When you connect to a session, the following message sometimes appears randomly:
  License was re-acquired.
• After you update the OS layer with May month Windows updates and publish to Provisioning Services, targets sometimes become unresponsive soon after sign-in. Affected machines then become inaccessible through RDP or console. As a workaround, republish with the previous OS layer.
• When the database is offline, SOAP server generates two error-level Windows events per minute. Periodic tasks trigger repeated entries and fill the event log while the database remains unavailable.
• Under heavy network load on XenServer, a Windows 10 x64 target might bugcheck with 0x50 PAGE_FAULT_IN_NONPAGED_AREA in the streaming driver send path.
• When AuditAction grows very large, Problem Report generation or audit export reads the whole table inefficiently. SQL Server CPU spikes, the console becomes unresponsive, and operations might run for hours. As a workaround, restart Soap Server and periodically export and remove old audit data.
• On XenServer 9.0.0 GA with Provisioning 2603, Streamed VM Setup Wizard might report no cache configurations on the PVS-Accelerator page despite a valid active configuration. This blocks accelerated streamed VM creation and displays a cache configuration error.
• When SoapServer processes scheduled work in a farm, a database deadlock might occur. You might see an SQL error that the transaction is the deadlock victim and needs rerun. This issue has no adverse effects.
• When you boot a VM with boot ISO, the display shows only target MAC address and omits PVS server information. This occurs whether or not a vDisk is assigned.
• After recent Windows updates on Server 2019 servers and targets, PVS performance counters, including vDisk read and write latency, might stop updating. As a workaround, run a tool such as typeperf to enable counters first.

Previously reported issues

• VMware targets with two vmxnet3 adapters might become unresponsive on a dark color screen during boot before Windows sign-in when the second adapter uses DHCP.
• On Windows Server 2025 targets, the client FQDN might not be sent to DHCP in some configurations. DHCP lease and DHCP-driven DNS records then do not update with the target FQDN, which might break DNS-dependent tools.
• When you configure boot devices with Boot Device Management, you cannot proceed beyond the Specify the Login Server page if you do all of the following:
  1. Select Use DNS to find the server.
  2. Select the Target Device is UEFI firmware checkbox.
  3. Enter an FQDN longer than 15 characters in the Server FQDN field.
  4. Clear the Target Device is UEFI firmware checkbox.
  5. Select Use static IP address for the Server. You can continue if you clear the FQDN field or reduce it to fewer than 15 characters.
• When targets are provisioned through an Azure Machine Catalog, you might not be able to sign in through Azure Portal Bastion in some cases, even when Citrix Workspace app access still works.
• When you use Manage in Virtual Apps and Desktops from an MCS-provisioned collection, the browser link might open a page that keeps loading instead of showing expected machine details.
• In Citrix Virtual Desktops Setup Wizard, selecting an existing Windows VM template might show a SnapIn is not responding message. You can dismiss the message, continue, and still create the device collection successfully.
• While moving Citrix Provisioning Servers between farms, moving the last server out of the old farm makes the old farm database inaccessible. You cannot use that old farm again. As a workaround, back up HKLM\\Software\\Citrix\\ProvisioningServices before moving the last server.
• Running Configuration Wizard in automated or answer-file mode (/a) might fail with an unhandled exception and crash in configurations where a service account value is empty.
• Target Device Update BDM dialog does not work if you select more than one NIC on the same Citrix Provisioning Server.
• When you switch a vDisk between Standard and Private modes, configured Max RAM size for cache might revert to the default value instead of retaining your previous value.
• In some cases, Configuration Wizard encounters an unexpected error and then closes without a message. You then need to check event logs, which often provide limited detail.
• When you use Configuration Wizard to create a new farm in an existing empty Azure SQL Database, unexpected errors make configuration fail. The log might report missing permission to alter database, missing database, or database state that blocks access checks. This occurs even when database state and permissions appear valid. If create-database permissions are a concern, use DbScript.exe.
• When you provision targets to Azure with an expired hosting connection client secret, BDM disk upload might fail with a long unclear error that does not identify the expired secret as the root cause.
• When you rerun Configuration Wizard on a Cloud-registered Provisioning server, Leave Citrix Cloud page might not appear in some flows. This can occur after selecting Create farm or Join existing farm, then going back and choosing Farm is already configured.
• P2PVS cannot create a vDisk when the source VM boots with UEFI. As a workaround, use Imaging Wizard. For example, run c:\\Program Files\\Citrix\\Provisioning Services\\ImagingWizard.exe P2vhdx win10 c:\\test to create VHDX file win10 in c:\\test. With UEFI imaging, you cannot specify individual volumes, imaging runs on the entire boot disk, and multiple disks are not supported.
• Selecting Get Template Properties while creating a device from a template might trigger an application error dialog. Dismissing the dialog can crash the PVS console.
• When you uninstall PVS Console and Server sequentially through the installer, a misleading warning might state that PVS Console is not detected even though both components are removed correctly.
• Upgrading PVS server from 2503 to 2507 CU1 might be interrupted in some cases and prevent successful completion of installation.
• In some MCS-PVS catalog configurations, VMs provisioned through MCS might be editable in the PVS console Properties dialog. Some changes, such as target device name, might not sync back to Delivery Controller.
• Under some MCS-PVS conditions, data returned by PVS PowerShell Snap-in might not match PVS database data. Some target-device information visible in the database might be missing from PowerShell query results.