Skip to content

Features and Enhancements in HyWorks v4.1-HF2

This document lists the new features and enhancements introduced in HyWorks v4.1-HF2.

HyWorks

New Sub-type for Nutanix Connector

Nutanix connector of type Prism Central now has the following sub-categories and configurations:

Sub-type Supported API Type & Recommendation
Prism Central (version >= 7.3) V4.0 and V4.2 (Recommended)
V3
Prism Central V>7.0 and <7.3 V4 (Recommended)
V3

Post-upgrade, if HyWorks controller detects an existing Nutanix provider that is eligible for the newer Sub-Type: Prism Central (version > 7.3) and API-Type: V4.0/4.2 (Recommended) configuration but is running on an older Sub-Type/API-Type, a Warning log will be generated for connector-related operations. This helps administrators identify providers that should be reconfigured to use Nutanix API v4.2.

  • Requirement: Applicable to Nutanix Prism Central v7.3 and above.

Important

  • Currently Open Issue with APIv4: Clone from snapshot feature does not work correctly with APIv4 (this issue is pre-existing with HyWorks v4.0).

    • Clone from snapshot (recovery point) with Nutanix Connector APIv4 does not use the selected recovery point and instead directly clones from the current VM state.

    • This issue has been identified in the Nutanix APIv4 and not in HyWorks. To mitigate this issue, APIv4.2 is integrated (Available from Nutanix PC v7.3).

  • Currently Open Issue with APIv4.2: APIv4.2, though provides support for the Clone from snapshot feature, but it does not provide an option to preserve MAC. Again this is identified as a problem with Nutanix APIv4.2.

    • To use MAC preserve, cloning must be done directly via the source VM instead of clone from snapshot.

  • For this release, these features are disabled as per the configured API version and Nutanix Sub-type. The hotfix for these issue(s) will be released soon.

Azure - Configurable Per-operation API Rate Limiting with Automatic Retry

HyWorks has an Azure connector with API throttling per operation type (read/write/delete/Graph/auth) and automatically retries upon receiving error responses from Azure (HTTP 429/503), preventing large provisioning or bulk teardown operations from failing under load. Concurrency limits are computed from admin-entered per-minute rate limits (Little's Law), and settings take effect without a service restart.

  • Requirement: Configured via new advanced settings

    • Enable Azure API rate limiting: (default On)

    • Azure API retry count

    • Azure API retry wait time (sec)

    • Azure read limit (requests/min)

    • Azure write limit (requests/min)

    • Azure delete limit (requests/min)

    • Microsoft Graph limit (requests/min)

    • Entra token limit (requests/min)

  • Remark: Feature is enabled by default; when disabled, behavior is identical to prior builds.

AVD - More detailed Operation Logs

Two AVD-related operation logs now include identifying details instead of a generic success message:

  • The host pool setup complete log now lists the actual names of the resource group, workspace, host pool, and application group created.

  • The application group user assignment log now lists the specific usernames assigned.

OpenStack - Improvised Connector Configuration and Provisioning Features

The OpenStack connector's VM provisioning flow has been improvised: image creation now uses a volume-snapshot-based (Glance/Cinder) flow, VM creation no longer requires pre-created network ports/NICs (passing a network ID now auto-creates them), and connector authentication is simplified to an application ID and application secret (a Project ID is no longer required).

  • Remark: Affects the "Add Connector" configuration and the creation of full-clone-from-snapshot VMs in OpenStack.

OpenStack - Capacity Plan and Scheduled Actions Support

Shared Hosted Desktop (SHD) pools created on the OpenStack connector can now be assigned a Capacity Plan, matching behavior already available for other providers/pool types.

Enable VM deletion from Connector for Static (non-deployed) Pools

Administrators can now choose to delete VMs directly from the hypervisor/connector when removing a non-deployed desktop pool created by adding pre-existing VMs from the connector or removing the desktop directly from the Desktops page.

  • Remarks:

    • The deletion is supported from the Desktops and Pools pages.

    • It does not delete computer objects from Active Directory (if desktops are domain-joined).

Restart Operation Uses Guest Reboot on Nutanix

Initiating Restart from the Management Console on Nutanix-backed desktops now performs a graceful guest OS reboot instead of a hard Reset, reducing the risk of data loss from an abrupt VM reset.

Connection Profile Updates for Request Redirection

The connection profile shared with the client now sends the appropriate virtual-channel parameters to Windows endpoints. Earlier, this had to be mentioned in the Connection Profile as an additional command-line parameter.

  • Remark:

    • Linux-based endpoints still require additional command-line parameters due to complex version dependencies.

Correlation IDs (CCID, GCID, VCID) added to ARS logs

ARS (Remote Syslog) and local Controller logs (EDController.txt) now include correlation IDs - CCID, GCID, VCID (and ACID) - across Login, Connect Desktop/App, Disconnect, Logout, idle/background session handling, and admin session operations, to make cross-referencing session events easier for customers such as JIO who correlate logs across systems.

  • Remark: Log message format is otherwise unchanged; only correlation ID fields are appended.

Restricted Management Console access on Session (Limited) nodes

On Controller nodes running in Session (Limited) mode, the Management Console now restricts visible/usable functionality to Dashboard, Licensing, Logs, and Syslog configuration only. Administrative and VDI management operations are no longer exposed on session-only nodes.

  • Remark: This is an access-control hardening change; it was introduced following an incident in which a session-only node was permitted to perform pool operations it should not have been.

Auto-fill of HySecure Token Signing Key

When an organization is configured to use HySecure Token authentication, the HySecure Token Signing Key is automatically populated with a default value, eliminating the need for the admin to retrieve it manually from the Gateway. Administrators can still override it with a custom key.

RDP Protocol Labels Updated in Connection Profile - Additional Command Line Tab

Protocol selection labels in the connection profile UI have been renamed from version-based names to OS-based names for clarity:

  • RDP 8.0 (Linux) > RDP (macOS) (As only MAC OS uses RDP 8.0)

  • RDP10 (Linux) > RDP (Linux) (As all Linux Endpoints now use the same RDP represented as RDP10)

  • The redundant RDP 7.0 (Linux) option has been removed (falls back to RDP (Linux)).

  • Remark: Cosmetic/usability change; no functional impact on session behavior.

Warning Shown before Power Operations on Inactive Pools

Attempting a Shutdown/Power Off (or similar power operation) on desktops belonging to an inactive pool now shows an explicit warning that the pool is inactive and the operation will not proceed, instead of the operation appearing to hang with no feedback.

APIs and Integrations (EDC.API / ARS)

Health API support on the Device Controller

A new Health API is available in the Device Controller to programmatically monitor node status.

  • Requirement: Reachable at the Controller node (primary/broker node in HA) over HTTPS on port 38866.

EDC.API and Controller changes to support "SMB Access Anywhere"

Backend changes to HyWorks and EDC.API to support the SMB Access Anywhere capability, including new Health and User Session APIs (GetLocalNodeHealth, GetUserSessionInfo).

  • Requirement: GetLocalNodeHealth uses an X-Api-Key; GetUserSessionInfo uses JWT Bearer (RS256) authentication with an API client registered in MMC (Issuer, Audience, RS256, matching public key, and "session-read" permission).

Additional fields in Create/Update basic reservation APIs

The Create/Update basic reservation APIs (EDC.API) now support specifying Floating vs. Personal VM assignment, a custom VM name prefix, and setting "Keep Desktop In Power On State" to Always on the pool created for the reservation.

Installer, Patcher, and Deployment Tools

Patcher preserves EDC.API configuration during patching

The HyWorks Patcher previously overwrote the installed EDC.API.exe.config with the shipped template, wiping customer-specific settings (connection strings, CertificateThumbprint, ManagementControllerUrl, OpenSearchPrefix, UseHttps, etc.). The patcher now merges the config instead of overwriting it: connection strings and existing customer values are preserved, new settings introduced by the patch are added, and removed settings are dropped. The original config is backed up before merging.

  • Remark: Eliminates the need for manual EDC.API reconfiguration after applying a patch.

Accops Hyper-V Manager (AHM) Log Configuration Improvement

AHM log files now roll over at a defined 50 MB per file / 10-file retention policy instead of growing unbounded. Routine health-check polling has been moved from INFO to DEBUG level, and health-status transitions are logged only when the state actually changes (Healthy/Unhealthy) rather than on every poll.

  • Remark: Reduces log volume and disk usage on AHM hosts; no functional change to health monitoring itself.