IMPORTANT: For clusters using VMware vSAN, you must first upgrade the vCenter Server system. Upgrading only the ESXi hosts is not supported.
Before an upgrade, always verify in the VMware Product Interoperability Matrix compatible upgrade paths from earlier versions of ESXi, vCenter Server and vSAN to the current version.
The typical way to apply patches to ESXi hosts is by using the VMware vSphere Update Manager. For details, see the About Installing and Administering VMware vSphere Update Manager.
ESXi hosts can be updated by manually downloading the patch ZIP file from VMware Customer Connect. From the Select a Product drop-down menu, select ESXi (Embedded and Installable) and from the Select a Version drop-down menu, select 6.7.0. Install VIBs by using the esxcli software vib update command. Additionally, you can update the system by using the image profile and the esxcli software profile update command.
The VMware Product Interoperability Matrix provides details about the compatibility of current and earlier versions of VMware vSphere components, including ESXi, VMware vCenter Server, and optional VMware products. Check the VMware Product Interoperability Matrix also for information about supported management and backup agents before you install ESXi or vCenter Server.
Virtual machines that are compatible with ESX 3.x and later (hardware version 4) are supported with ESXi 6.7 Update 2. Virtual machines that are compatible with ESX 2.x and later (hardware version 3) are not supported. To use such virtual machines on ESXi 6.7 Update 2, upgrade the virtual machine compatibility. See the ESXi Upgrade documentation.
VMware introduces a new Configuration Maximums tool to help you plan your vSphere deployments. Use this tool to view VMware-recommended limits for virtual machines, ESXi, vCenter Server, vSAN, networking, and so on. You can also compare limits for two or more product releases. The VMware Configuration Maximums tool is best viewed on larger format devices such as desktops and laptops.
For information about upgrading with third-party customizations, see the ESXi Upgrade documentation. For information about using Image Builder to make a custom ISO, see the ESXi Installation and Setup documentation.
IMPORTANT: Before an upgrade, always verify in the VMware Product Interoperability Matrix compatible upgrade paths from earlier versions of ESXi and vCenter Server to the current version.
Disclaimer
The bulletin listing in these release notes is provided for informational purposes only. This listing is subject to change without notice and the final list of released patch bundles will be posted at:
THIS LISTING IS PROVIDED "AS-IS" AND VMWARE SPECIFICALLY DISCLAIMS ALL REPRESENTATIONS AND WARRANTIES, EXPRESS OR IMPLIED, INCLUDING ITS MERCHANTABILITY, NONINFRINGEMENT, AND FITNESS FOR A PARTICULAR PURPOSE. VMWARE DOES NOT REPRESENT OR WARRANT THAT THE LISTING IS FREE FROM ERRORS. TO THE MAXIMUM EXTENT OF THE LAW, VMWARE IS NOT LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, OR CONSEQUENTIAL DAMAGES, HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, EVEN IF VMWARE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
If the USB device attached to an ESXi host has descriptors that are not compliant with the standard USB specifications, the virtual USB stack might fail to pass through a USB device into a virtual machine. As a result, virtual machines become unresponsive, and you must either power off the virtual machine by using an ESXCLI command, or restart the ESXi host.
When attempting to set a value to false for an advanced option parameter in a host profile, the user interface creates a non-empty string value. Values that are not empty are interpreted as true and the advanced option parameter receives a true value in the host profile.
A rare race condition with static map initialization might cause ESXi hosts to temporarily lose connectivity to vCenter Server after the vSphere Replication appliance powers on. However, the hostd service automatically restarts and the ESXi hosts restore connectivity.
In rare cases, an empty or unset property of a VIM API data array might cause the hostd service to fail. As a result, ESXi hosts lose connectivity to vCenter Server and you must manually reconnect the hosts.
ESXi hosts might intermittently fail with a purple diagnostic screen with an error such as @BlueScreen: VERIFY bora/vmkernel/sched/cpusched.c that suggests a preemption anomaly. However, the VMkernel preemption anomaly detection logic might fail to identify the correct kernel context and show a warning for the wrong context.
This issue is resolved in this release. The fix covers corner cases that might cause a preemption anomaly, such as when a system world exits while holding a spinlock. The preemption anomaly detection warning remains @BlueScreen: VERIFY bora/vmkernel/sched/cpusched.c, but in the correct context.
If you change the DiskMaxIOSize advanced config option to a lower value, I/Os with large block sizes might get incorrectly split and queue at the PSA path. As a result, ESXi hosts I/O operations might time out and fail.
In rare cases, some Intel CPUs might fail to forward #DB traps and if a timer interrupt happens during the Windows system call, virtual machine might triple fault. In the vmware.log file, you see an error such as msg.monitorEvent.tripleFault. This issue is not consistent and virtual machines that consistently triple fault or triple fault after boot are impacted by another issue.
This issue is resolved in this release. The fix forwards all #DB traps from CPUs into the guest operating system, except when the DB trap comes from a debugger that is attached to the virtual machine.
If you plug in and out a physical NIC in your vCenter Sever system, after the uplink is restored, you still see an alarm in the vSphere Client or the vSphere Web Client that a NIC link on some ESXi hosts is down. The VOBD daemon might not create the event esx.clear.net.redundancy.restored to remove such alarms, which causes the issue.
In some cases, when multiple USB or SD devices with different file systems are connected to your vCenter Server system, ESXi installation by using a ks.cfg installation script and a USB or SD booting device might fail.
In the first minutes after mounting a VMFS6 volume, you might see higher than configured rate of automatic unmap operations. For example, if you set the unmap bandwidth to 2000 MB/s, you see unmap operations running at more than 3000 MB/s.
For independent nonpersistent mode virtual disks, all writes are stored in a redo log, which is a temporary file with extension .REDO_XXXXXX in the VM directory. However, a reset of a virtual machine does not clear the old redo log on such disks.
Rarely, in certain configurations, the shutdown or reboot of ESXi hosts might stop at the step Shutting down device drivers for a long time, in the order of 20 minutes, but the operation eventually completes.
The virtual network adapter VMXNET Generation 3 (VMXNET3) uses buffers to process rx packets. Such buffers are either pre-pinned or pinned and mapped during runtime. In rare occasions, some buffers might not un-pin and get re-pinned later, resulting in a higher-than-expected pin count of such a buffer. As a result, virtual machines might fail with an error such as MONITOR PANIC: vmk: vcpu-0:Too many sticky pages..
Due to a rare lock rank violation in the vSphere Replication I/O filter, some ESXi hosts might fail with a purple diagnostic screen when vSphere Replication is enabled. You see an error such as VERIFY bora/vmkernel/main/bh.c:978 on the screen.
In vCenter Server advanced performance charts, you see an increasing number of packet drop count for all virtual machines that have NetX redirection enabled. However, if you disable NetX redirection, the count becomes 0.
In particular circumstances, when you power on a virtual machine with a corrupted VMware Tools manifest file, the hostd service might fail. As a result, the ESXi host becomes unresponsive. You can backtrace the issue in the hostd dump file that is usually generated in such cases at the time a VM is powered on.
After a reboot of an ESXi host, encrypted virtual machines might not auto power on even when Autostart is configured with the Start delay option to set a specific start time of the host. The issue affects only encrypted virtual machines due to a delay in the distribution of keys from standard key providers.
When the SLP service is disabled to prevent potential security vulnerabilities, the sfcbd-watchdog service might remain enabled and cause compliance check failures when you perform updates by using a host profile.
ESXi updates and upgrades might fail on Mac Pro servers with a purple diagnostic screen and a message such as PSOD - NOT IMPLEMENTED bora/vmkernel/hardware/pci/bridge.c:372 PCPU0:2097152/bootstrap. The issue occurs when Mac Pro BIOS does not assign resources to some devices.
Some Linux kernels add the IPv6 Tunnel Encapsulation Limit option to IPv6 tunnel packets as described in the RFC 2473, par. 5.1. As a result, IPv6 tunnel packets are dropped in traffic between virtual machines with ENS enabled, because of the IPv6 extension header.
Resource allocation for a delta disk to create a snapshot of a large virtual machine, with virtual disks equal to or exceeding 1TB, on large VMFS6 datastores of 30TB or more, might take significant time. As a result, the virtual machine might temporarily lose connectivity. The issue affects primarily VMFS6 filesystems.
In case of temporary connectivity issues, ESXi hosts might not discover devices configured with the VMW_SATP_INV plug-in after connectivity restores, because SCSI commands that fail during the device discovery stage cause an out-of-memory condition for the plug-in.
If you import a virtual machine with a virtual USB device that is not supported by vCenter Server, such as a virtual USB camera, to a vCenter Server system, the VM might fail to power on. The start of such virtual machines fails with an error message similar to: PANIC: Unexpected signal: 11. The issue affects mostly VM import operations from VMware Fusion or VMware Workstation systems, which support a wide range of virtual USB devices.
795a8134c1