It’s been a busy week of product updates and improvements. Here is what has been released last week.

AOS (STS) 6.0.2.2

AOS integrates storage, compute and networking resources into a scalable, secure, resilient and easy-to-use software-defined infrastructure.

The release resolves the following critical issue:

Prism Gateway

• Resolved an issue where attempts to upgrade to AOS 6.0.2 might stall as the Controller VM eth0 network interface is not available after the Controller VM restarts as part of the upgrade process.

AOS (LTS) 5.20.2.1

The release resolves the following critical issues:

Cluster Conversion

• Improved network validation to ensure that the networks that do not have any VMs attached are, by default, retained after an ESXi to AHV cluster conversion. Contact Nutanix Support if you do not wish to retain networks without any VMs.
• Resolved an issue where during cluster conversion the mapping of the target networks is not properly set to the default virtual switch vs0.
• Improved the cluster conversion process by running 20 VM conversions in parallel.

Data Replication / Protection

• The retry or backoff logic applied while waiting for stretch data interface resolution is improved to reduce latency and stun times for quiesced snapshots.

Flow

• Resolved an issue where stale cache, populated when VM migration happens while Flow is disabled, results in incorrect rules to be programmed. For further details, see KB 12200.

Microservices Platform (MSP) 2.3.4

MSP is a managed platform based on Kubernetes for managing containerized services running on PC. The lifecycle of MSP is controlled by a service running on Prism Central called MSP Controller. MSP supports various containerized services like IAMv2, ANC and Objects and more services will be on-boarded in future. You must upgrade MSP before upgrading any of these services.

Following are the resolved issues in this release:

• Fixed the issue where Prism Central was not able to reach corp DNS server due to 127.0.0.1 entry in the /etc/resolv.conf file restricting the user to configure FQDN on Prism Central.
• Fixed the issue where external nameserver were erased from Object VMs post reboot. This could have caused connectivity issues to LDAP, external FQDN, and so on.
• Fixed the issue where MSP Controller upgrade through LCM darksite bundle failed; however, in the user interface upgrade was seen as successful.
• Fixed the issue where object data path had seen issue in case of eth0 network down or up case on Objects cluster running 1.0.8 or below versions.

Life Cycle Manager Version 2.4.3.3

The Life Cycle Manager (LCM) tracks and upgrades both software and firmware versions of all entities in the cluster, completely in-band with Nutanix Prism Element and Prism Central.

New Features:

Change path for sudo command

• For all references to sudo, LCM 2.4.3.3 replaces the path /usr/bin/sudo with sudo

Resolved Issues:

ESXi precheck failing with UEFI

• The precheck test_esxi_bootbank_space verifies that the system has enough space to restart into the Phoenix ISO. Foundation 4.4.1 added the ability to mitigate low bootbank space, so systems that use Foundation 4.4.1 or later disable this precheck. However, if the system uses UEFI, Foundation's mitigation does not work. LCM now enables or disables test_esxi_bootbank_space depending on the detected boot mode.

Inventory failing on Intel

• Fixed an issue on Intel systems where inventory sometimes failed with the error Unable to get the fru details. Failed to run payload_info cmd.

ESXi systems returning Failed to power on VMs error after update

• Fixed an issue where LCM treated vSphere cluster services (vCLS) VMs as pinned VMs that LCM needed to restart after an update. ESXi automatically restarts vCLS VMs after a host restart, so LCM was trying to start VMs that had already started, resulting in the observed error.

Be sure to share your upgrade experience in the community forums.