The failover functionality on Windows Server 2022 works correctly.

Changes were made to prevent database performance issues in large environments.

If SQL Server stops a SQL process, LA database maintenance logs the issue and continues running instead of waiting indefinitely on the process.

When devices include duplicate data, inventory jobs no longer fail with the message Violation of PRIMARY KEY constraint 'PK_NCM_Entity_Logical'.

Additional ciphers are supported for collecting data via SSH. This fixes the connectivity issues with certain versions of the FortiGate software. For the list of supported ciphers, see Supported algorithms and ciphers.

If NCM still cannot connect to Fortigate devices after the upgrade, complete the additional steps in this article.

Having duplicate names for the same node ID no longer causes an upgrade to fail.

Upgrades no longer fail with the message Error while executing script - ALTER TABLE DROP COLUMN failed because column 'InterfaceID' does not exist in table 'NCM_ARPTables'.

Config change approval requests are no longer deleted because of account limitations.

If the CVE field on the Firmware Vulnerabilities Settings page contains the CPE match data feed URL instead, a message is displayed and the settings cannot be saved.

01490187, 01398440

The Collector service no longer reports errors related to SwitchStack UNIQUE KEY constraints, similar to:

ERROR SolarWinds.Collector.DataSynchronization.InventoryManager - SyncEntity for SSM:53 failed: System.Data.SqlClient.SqlException (0x80131904): Violation of UNIQUE KEY constraint 'UK_NPM_SwitchStackMember'. Cannot insert duplicate key in object 'dbo.NPM_SwitchStackMember'. The duplicate key value is (965, FOC2327T1DZ)

01516014, 01429174

The automatic refresh of an expired API key for Palo Alto devices has been handled.

Interface traffic for VeloCloud Networks is polled correctly.

The common MIB2 ifHCxxoctets counters do not work consistently for VeloCloud Edge devices when the device is configured as an HA (high availability) pair and its current state is StandBy. To overcome this issue, SolarWinds implemented a custom VeloCloud interface traffic data SNMP poller for getting OutTotalBytes/InTotalBytes counters from a private VeloCloud MIB.

When retrieving information about Meraki SD-WAN tunnels, all VPN statuses for an organization are retrieved, even when there are more than 300.

A new MAC property for Device Studio Node Details pollers allows specification of the data source (OID) for MAC address collection. With this change, custom node detail pollers that were created in 2023.4 or earlier work correctly, and you can set the MAC data source for them.

Empty DNS field in the DB NodesData table for Viptela, VeloCloud, and Aruba orchestrators no longer causes polling to fail, and an error is no longer displayed similar to:

ERROR SolarWinds.Orion.Pollers.Framework.PollerBase - Failed to poll Viptela Orchestrator with Node ID=7: System.ArgumentException: ipAddressOrHostname

In Centralized Settings, the new setting SendBgpInfoRequests has the following values:

• True: Sends requests to retrieve BGP Annoucement and WHOIS network data.
• False: Stops sending these requests. Only stored data is used.

Use this setting to temporarily stop requests if the NetPath_Networks table is filled during BGP polling and you see errors such as:

ERROR SolarWinds.NetPath.Bgp.RawClients.RawClient - (null) [Flow 40442] TcpQueryAsync have failed. Endpoint: Unspecified/whois.iana.org:43. Exception: System.OutOfMemoryException: Exception of type 'System.OutOfMemoryException' was thrown.

N/A

The Wireless Heat Map editor displays the owner information correctly.

Wireless heat map generation no longer fails when some access points return invalid data. The heat map generates, but problematic access points are ignored.

The performance of SD-WAN charts on the SD-WAN Orchestrator Details view has been improved.

If a custom property has a NULL value for some nodes and an empty string for other nodes, polling with a Universal Device Poller (UnDP) and then sorting by the custom property no longer returns an error message.

When you edit a group, saving the Add & Remove SolarWinds Platform Objects page without making any changes no longer removes any existing objects.

Detection of L3 Next Hop connections for VPN tunnels has been improved. Previously, having a full L3 Interface-to-Interface connection required that both nodes reported valid Next Hop addresses for each other. Now, only one node needs to report the connection. This default behavior can be changed with the advanced configuration option ResolveTargetInterfaceForNextHop.

A server without an IP address does not cause the List of Virtual Servers widget to display an exception.

SolarWinds IP address and port pollers no longer return misleading results for Meraki appliance models that have a fallback cellular uplink.

Queries associated with the Duplex Mismatch alert have been tuned to improve the performance of the SolarWinds Platform Web Console.

If one IP address has more than RemoteSystemName in LLDP data, inaccurate connections are no longer reported in the topology table.

In previous versions, even though a user had the right to manage nodes and could add new devices or edit existing ones, the user was not able to add an interface through the NetFlow Events resource. Now, users with Node Management rights are able to add interfaces. Admin rights are no longer required for this task.

During on-demand DNS resolution, when an option to resolve IP addresses is disabled and then reenabled after maintenance, IP addresses are resolved and the Domains and Resources widgets display the correct information.

Creating and managing an application with a large port range is handled properly in NTA.

Netflow PDU is marked as malformed when octets or packets have a negative value, fewer octets than packets are reported, and maximum transition unit (MTU) is larger than 64. These types of invalid data are detected and discarded.

01567985

When you enable the High Receive Percent Utilization with Top Talkers out-of-the-box alert or High Transmit Percent Utilization with Top Talkers out-of-the-box alert and configure an alert action to send a PDF, the PDF file is generated correctly and it no longer contains an error similar to:

Unexpected Website Error Failure parsing NetObject key.

An API poller that includes a null value no longer causes sent requests to fail with the message Bad Request.

When Windows Update Monitoring fails to connect to a remote server, the correct message is displayed instead of a message saying that initial polling is in progress.

An API poller can be removed using the SWIS SDK. The Delete command no longer returns an error.

Asset Inventory returns data from nodes running a Linux-based operating system.

When an Exchange user's password includes a special character, AppInsight for Exchange displays correct data for Custom PowerShell Monitor components, rather than showing them in an unknown state.

A VM in an Azure account is polled correctly, and the status is not shown as unknown.

A SolarWinds query to retrieve component statuses was tuned to run more quickly.

N/A

The performance of the Last 10 Events widget has been improved by adding indexing to the Orion.Events table.

N/A

Non-admin users have access to the new Azure dashboard.

Logging has been improved to provide information about the time spent enumerating files and folders.

Running Asset Inventory on a node with the Linux agent installed no longer fails with errors such as the following:

Failed to parse or validate SNMP parameters provided by user

If the upgrade of an Oracle ODBC driver causes problems with the SolarWinds Platform Installer, the following entry can be entered in the installer's registry key to avoid running the Oracle ODBC driver installation executable:

SkipOracleOdbcDriver = True

If you add a node polled with WMI and you do not include the domain with the credentials, asset inventory polling works correctly and the Windows Update Information resource does not display errors such as:

The following error with errorcode 0x8009030d occurred while using Negotiate authentication: A specified logon session does not exist.

Ports that are administratively shut down are not imported if this option is deselected in the discovery.

The MAC addresses of devices connected to Fortiswitches are displayed correctly.

The version of OpenJDK shipped with the product has been upgraded to 11.0.22.

Nutanix Hardware Health polling works in large environments.

Free nodes are no longer included in license consumption when adding a new virtual environment.

VMAN polling now collects reboot values for hosts during the sample collection (every 10 minutes), which enables the Host Rebooted alert to work correctly.

The Node Details widget for aVMware ESX server shows the correct number of CPUs.

01283241

Nutanix Controller VM (CVM) credentials can be added to Nutanix Clusters.

N/A

The Save new credentials button is disabled during credentials validation, so you do not run the same verification multiple times.

Database maintenance no longer fails with the following message:

Violation of PRIMARY KEY constraint 'PK_VoipOperationStdDevResults_Hourly'. Cannot insert duplicate key in object 'dbo.VoipOperationStdDevResults_Hourly'.

Polling Cisco CallManagers with attached gateways does not increase CPU usage or produce an error.

The Business Layer on an additional polling engine no longer fails to start with the following message:

BusinessLayer.Plugin (null) - IP SLA Monitor service [SW550SLRWPA037] failed to start.System.ServiceModel.FaultException1[System.ServiceModel.ExceptionDetail]: An item with the same key has already been added.

01589368

The VNQM business layer on the additional polling engine no longer times out when receiving license information from the main polling engine in environments with many nodes.

CDR data is processed when columns contain Cisco-specific call termination cause calls, and an error is no longer displayed, similar to:

ERROR BusinessLayer.Task.CdrCmrCsvFileReader (null) - Cannot convert int value [column:destCause_value,rawValue:2751463455]“. As a result, this call data is not saved to DB.

Also, the VoIP Call Details resource is rendered correctly on the Call Details page, and the VoIP Search page opens without errors.

An issue that could cause menu failures or prevent IP SLA from loading has been fixed.

SolarWinds no longer displays invalid messages saying that the license limit is exceeded. This was caused by attempts to verify licenses before they were loaded.

The Last Polled field on the Call Manager page now displays the time in the time zone of the server, not in the UTC time zone.

Juniper RPM operations are polled and the status changes correctly.

When an Avaya call manager has a large number of phones, gateways, and other attachments (for example, over 1,000), polling is completed during one polling cycle.

Integration with third-party FTP server software has been improved. VNQM can download call manager data from FTP servers such as FileZilla and others.

01561273

IP SLA UDP Jitter operations are no longer deleted from VNQM if it is created with an external node as a target.