The CVE Node Matching process no longer fails with an error about a missing value for the 'Value' parameter.

The CVE Data Import no longer fails when run from an Azure SQL DB.

When the Job Engine executes the RemoveEmptyRouters method on an agent, it no longer returns the error FileNotFound Orion.Core.Common.dll.

The Intelligent Map widget displays the Map name format field without requiring the user to clear the browser cache.

To improve the performance of world maps, the following changes have been implemented:

• Elements that do not provide value (such as labels without titles or without coordinates) are automatically removed.
• A health check warns you if multiple labels are assigned to the same coordinates.

Editing a map created from a device works as expected.

A user whose account name includes an apostrophe can view Intelligent Maps.

On Intelligent Maps, the colors of the Shutdown, External, Unknown, Unpluggable, and Unreachable status icons are no longer the same.

When an object on an Intelligent Map has label text defined by macros, the label text is not displayed when the map is opened on a modern dashboard.

When live updates are on for an Intelligent Map, the map is now updated when interface statuses or statistics change.

If the SolarWinds Platform database is hosted on an Azure SQL DB or Azure SQL Managed Instance, the Configuration Wizard no longer fails with the error message Some of your modules (Platform Connect) don't support Azure database.

The SolarWinds Platform is compliant with .NET 4.0 STIG.

When you log in to the SolarWinds Platform Web Console through an additional web server and create or edit a map, audit map messages show the correct account ID instead of user SYSTEM.

Downloading the Linux agent offline installer no longer fails when the system locale of the SolarWinds Platform server is German.

When IPv6 is disabled on a polling engine and the SolarWinds Agent is distributed through a Golden Master image, the agent configuration no longer contains the IPv6 loopback ::1.

If you display the Planned Alert Suppression column on the Managed Nodes page and then perform a search, the following error message is no longer displayed:

Server error: undefined
An error has occurred.

New custom properties cannot be named OrionGroupName. In 2024.4 and later versions, OrionGroupName is reserved. For a list of all reserved terms, see this article.

The List Resources view is displayed correctly for nodes with agent polling.

When you select the New Diagnostics option, the Collect new diagnostics dialog lists nodes as well as polling engines.

Data remaining from evaluations of earlier versions of NTA no longer prevents upgrading to current versions of SolarWinds Platform products.

When an additional polling engine installation is blocked because a centralized upgrade was started but not completed, the correct message is displayed.

Repairing the SolarWinds Administration Service (SWA) from the Control Panel no longer fails with an error because the required .msi file is missing.

Leaving a modern dashboard open without activity long enough for a session timeout does not result in the following error:

Server Error in '/apps' Application.
Value cannot be null.
Parameter name: sessionId

The Configuration Wizard works correctly when "Reconfigure database connection only" is selected.

When a user upgrades from a legacy version of the Orion Platform that used Microsoft SQL 2014 or earlier as the database, database maintenance no longer fails with messages that database columns cannot be modified.

In the ServiceNow integration, the Impact drop-down is populated with the correct options instead of only None and High.

On the Manage Nodes page, the Group by option previously labeled "Status" is now labeled “Group by Node”. This change is to clarify that when Interfaces are selected, grouping by status groups the interfaces based on the status of the associated node.

The database was tuned to prevent problems with alerting.

When one or more nodes that are part of an anomaly-based alert become unmanaged, the alert works as expected.

Editing an anomaly-based alert and changing aspects other than the selected entities no longer results in issues such as the alert details page being empty or the previously selected entities no longer being selected.

When nodes related through topology trigger alerts, the AlertStack no longer fails to create a cluster with errors such as:

ERROR SolarWinds.AIIM.Issues.BusinessLayer.IssueBuilder.AlertNotificationProcessor - Process Alert Notification Failed System.Collections.Generic.KeyNotFoundException: The given key was not present in the dictionary.

When the SolarWinds Administration Service (SWA) uses a Powershell function to retrieve the hash value of a product catalog, it no longer returns an invalid value in some situations.

The custom property editor no longer treats null and empty as different values.

If the same node is added twice, with the first instance polled via WMI and the second by an agent, you can no longer edit the first instance and change the polling method to agent. This would cause polling to fail.

When you are adding entities to a Perfstack chart and the list of entities loads slowly, the Add Entities dialog is no longer missing checkboxes, sorting options, and filters.

Database maintenance no longer removes all information about a triggered global alert from the database.

Users no longer require "Alert Management" rights (or higher-level Admin rights) to receive alert notifications in the Desktop Notification Tool. The "Allow Account to Clear/Acknowledge messages" permission is sufficient.

On the All Active Alerts page, alerts can be sorted by the ServiceNow Incident Number column or the Assigned to column.

When a node is scheduled for maintenance mode in the future but not currently in maintenance mode, the drop-down command menu displays "Cancel Maintenance" instead of "Resume Alerts (Unmute)".

If a node is assigned to an additional polling engine (APE) and the APE becomes unreachable, users can still edit the value of custom properties on the node.

When you select specific nodes on the Manage Nodes page and choose Export Custom Property Values, the export file contains only the custom property values for the selected nodes, not values for all nodes.

After an unsuccessful connection, active agents will wait approximately 60 seconds before another connection attempt, which reduces the load on a polling engine when a lot of active agents are trying to reconnect after a maintenance window.

When you select multiple nodes and schedule a maintenance window, the time of the window is displayed correctly.

When "Send a GET or Post Request to a Web Server" is selected as an alert action and the HTTP post body is in JSON format, line breaks in variables no longer prevent the JSON from being parsed correctly, resulting in an error.

When a Cisco Catalyst 8000 or Catalyst 8300 series device is monitored, the Machine Type includes the model number instead of identifying it only as Cisco.

When an Arista device is monitored, the Machine Type correctly identifies the vendor model number instead of identifying it as Arista Networks.

The Management widgets on the Node Details and Volume Details pages no longer disregard custom language packs.

The SolarWinds Platform no longer reports incorrect CPU usage for some Cisco devices.

The SolarWinds Platform has been updated to more accurately report CPU usage on computers running Windows 8 and later.

When a monitored node's IP address changes, the node is no longer assigned to the wrong agent.

Reports now show data for the selected time period.

When a limitation is applied to an account, modern dashboards are displayed correctly for users who log in with that account.

When the trap service receives a trap with a varbind that contains an OID with the value 0.0, it accepts the value as null. It no longer tries to query the sending device for the value.

The log management portion of database maintenance performance was improved to prevent failures due to timeouts.

The SSH button no longer resolves to unsafe:ssh://10.11.11.101 in modern dashboards.

*Installations and upgrades no longer fail with the error Cannot get packages from the main poller. More data is available.

Firmware vulnerability updates correctly identify the release numbers of Juniper Junos devices.

The Last Config Backup Date widget on the Config Summary view no longer shows a different value than the Last Backup Date on the Configuration Management view in certain situations. Both values reflect the last attempted download time.

To save job results to a non-default location, you must provide credentials to give NCM write access. If you change only the default file name and then revoke write access, the following error message no longer appears when you edit the job:

Exception has been thrown by the target of an invocation.

NCM made improvements to prevent slow performance in the web console.

Out-of-the-box firmware upgrade templates and config change templates are no longer duplicated each time a user upgrades or re-runs the Configuration Wizard.

When an inventory scan runs, neither of the following errors are written to the C:\ProgramData\SolarWinds\Collector\Logs\Collector.Service.txt file when NCM writes data to the NCM_CiscoFlash table:

System.Data.SqlClient.SqlException (0x80131904): Cannot insert the value NULL into column 'FlashFileName', table 'GNO_Orion2.dbo.NCM_Cisco_FlashFiles'; column does not allow nulls. INSERT fails.

System.Data.SqlClient.SqlException (0x80131904): Arithmetic overflow error converting expression to data type int.
The statement has been terminated.

If a config change report specifies that an email should be sent only when there are changes, and the job runs but there are no changes, the report no longer generates the message Emailed report (when no email was sent). The report now generates the message No changes were detected. An email will not be sent.

The SolarWinds SCP server, which is installed with NCM, now has DSA, RSA, and ECDSA public key lengths that are long enough to be considered secure.

NCM monitors for Data Processing Engine (DPE) jobs that are hung in the post-processing phase. If it detects a non-responsive job, it stops the job, clears any related artifacts from the cache, and retries the job. If needed, advanced options to adjust default behavior are available to SolarWinds Support.

*Adding the macro ${N=Alerting;M=Notes} to an alert that includes an NCM action correctly includes notes from the alert in the alert notification.