VMware NSX-T Data Center 3.2.3.1 Release Notes

VMware NSX 3.2.3.1 | 27 JULY 2023 | Build 22104592 Check for additions and updates to these release notes.

What's New

NSX-T Data Center 3.2.3.1 is an update release that comprises bug fixes only. See "Resolved Issues" below for the list of issues resolved in this release. See the VMware NSX-T Data Center 3.2 Release Notes for the list of new features introduced in NSX-T 3.2. See the VMware NSX-T Data Center 3.2.3 Release Notes for the current known issues.

Compatibility and System Requirements

For compatibility and system requirements information, see the VMware Product Interoperability Matrices and the NSX-T Data Center Installation Guide.

Upgrade Notes for This Release

For instructions about upgrading the NSX-T Data Center components, see the NSX-T Data Center Upgrade Guide.

Available Languages

NSX-T Data Center has been localized into multiple languages: English, German, French, Italian, Japanese, Simplified Chinese, Korean, Traditional Chinese, Italian, and Spanish. Because NSX-T Data Center localization utilizes the browser language settings, ensure that your settings match the desired language.

Document Revision History

July 25, 2023. First edition.

Resolved Issues

• Fixed Issue 3244630: nestdb process on the ESXi host may fail to restart due to ramdisk being full.

  In some circumstances when the NestDB process on the ESXi host crashes and restarts, and if nestdb determines it is in an inconsistent state, it makes a copy of the current db and performs a resync with the NSX controller process. During this operation, the database files are written into a "lost" sub-folder and after multiple crashes cause "No space left on device", leading nestdb in a down state. This fix addresses the “lost” sub-folder creation process.

• Fixed Issue 3210231: LB Access log Printing nginx config under syslog.

  Access log memory is not cleared before the access log info is filled. LB configuration is leaked into the syslog. There is no impact on LB service.

• Fixed Issue 3210233: Tier1 state API delayed up to 50 seconds.

  GET /policy/api/v1/infra/tier-1s/<Tier1_name>/state API response delayed up to 50 seconds.

• Fixed Issue 3214571: Rules within nested groups may not work as expected on hosts in some cases.

  DFW rules are not getting programmed correctly under the VM's DFW filter, resulting in a traffic drop.

• Fixed Issue 3214569: Controller service is down in NSX-T Managers with ccp_oom core dumps.

  NCP clients can create orphan entries in the firewall table leading to OOM conditions.

• Fixed Issue 3206290: Apache benchmark packets are dropped when snat is configured.

  When the TCP server reuses the port from the TCP TIME_WAIT state flows and the client starts a new TCP session with a window scaling option, the packets from the server are dropped when they are bigger than a certain size.

• Fixed Issue 3245898: Core generated for the datapath during the Edge Node crash and failover.

  Edge Node crashes when a failover occurs and a DNS request packet is received on an existing state before the expiry of the connection. Process restart recovers the system but there is some loss of traffic in between.

• Fixed Issue 3239944: DHCP server IP getch met issue "cgw-dhcp-server-lp Error:null".

  After the upgrade of NSX-T to a higher version, the DHCP relay IP in Proton does not match the DHCP server IP in the Policy on the Tier-1 gateway.

• Fixed Issue 3238422: NSX IDPS engine is crashing in ESXi and generating core dumps.

  IDPS engine down alarms are observed on the MP dashboard and cores are seen on the ESXi host.

• Fixed Issue 3238430: Unable to connect Local Manager to Global Manager after replacing the Certificates for Local Managers and VIP.

  The GET /api/v1/trust-management/site-certificates API stopped working and always returned an error.

• Fixed Issue 3237011: Bond0 config lost on Edge Bare Metal post reboot.

  During Edge bootup, renaming of the network interface may fail due to a 'device already in use' error. This may lead to problems with the bond management interface.

• Fixed Issue 3238358: Bond Management interface breaks after upgrade from version 3.2.1 to 3.2.3.

  When a Bond Management interface is configured, NSX Edge 3.2.1 and lower versions have a udev rules file that contains entries with the same MAC address for different interfaces. However, after upgrading from 3.2.1 to a higher version, the secondary bond slave interface in the /etc/network/interfaces file does not have the ‘pre-up sleep 5’ stanza. This causes the bond MAC address to be inconsistent when the Edge is rebooted.

• Fixed Issue 3237007: PSOD occurs in ESXi Server.

  A PSOD occurs when the port that is used in the LTA session is disabled while the session is still active.

• Fixed Issue 3236433: Log messages are not reaching the logging server.

  If you use FQDN to configure a logging server, it can fail if more than one IP address is returned with the FQDN lookup.

• Fixed Issue 3236992: vMotion SVMs cause traffic flows to fail.

  After the vMotion of SVM, packet drop has been observed in existing traffic flows.

• Fixed Issue 3237229: VNI last sync time is getting updated when there is no change in any other field.

  Prior to 3.2, the Virtual Network Interface (VNI) object of the VM is updated only if there is any change in the properties of the object. However, from 3.2, when synchronization with the host occurs, even if there is no change in the network interface, last_sync_time is updated. Due to these updates, even if there is no change, there is a performance impact.

• Fixed Issue 3236430: After deletion, the entries in PortDiscoveryProfileBindingMap are not properly cleared.

  There could be memory exhaustion, leading to java.lang.OutOfMemoryError, due to the stale PortBindingMaps and their associated GenericPolicyRealizedResource entities.

• Fixed Issue 3236435: vMotion SVM cause traffic flow to fail or drop.

  TCP flows that are passing over an SVM fail if the SVM is moved using vMotion.

• Fixed Issue 3236374: Intermittent Packet loss during NSX-V to NSX-T migration at the edge cutover.

  Intermittent traffic loss at the edge cutover during the NSX-V to NSX-T migration for VRA routed networks going through downlink cutover.