Virtuozzo Hybrid Infrastructure 7.2 (7.2.0-246)¶
Issue date: 2026-01-20
Applies to: Virtuozzo Hybrid Infrastructure 7.2
Virtuozzo Advisory ID: VZA-2026-001
1. Overview¶
In this release, Virtuozzo Hybrid Infrastructure introduces support for two-factor authentication (2FA) for system administrators and self-service users, along with several other new features and improvements. Additionally, this release delivers stability fixes and addresses issues found in previous releases.
2. New Features¶
[Security] System administrators and self-service users can now configure TOTP-based two-factor authentication (2FA) using Google Authenticator or Microsoft Authenticator. Administrators can enforce 2FA usage to strengthen compliance and improve account security.
[Compute service] Custom MAC address prefixes can now be set for all VM network interfaces. This allows system administrators to avoid MAC duplication when running multiple clusters within the same L2 network segments.
[Compute service] Support for new guest operating systems:
Red Hat Enterprise Linux 10
AlmaLinux 10
Rocky Linux 10
SUSE Linux Enterprise 15 SP6 and SP7
Oracle Linux 10
3. Improvements¶
[Updates] QEMU hypervisor updates no longer require rolling maintenance with VM migration and node reboots. Future updates to the compute service that include the QEMU hypervisor will not trigger rolling maintenance, reducing operational overhead.
[Core storage] Sequential write performance doubles on tiers using replication redundancy with checksumming enabled. This improvement benefits workloads such as append-only databases and speeds up mass volume backups and snapshot operations.
[Core storage] Improved erasure coding schema selection when managing storage policies. The system now recommends an optimal schema based on the number of available failure domains. Administrators can also configure custom M+N schemas for non-standard configurations directly in the admin panel, without using command-line tools.
[Object storage] A new S3 health check API endpoint allows system administrators to verify node health from an external HAProxy load balancer, ensuring traffic is routed only to healthy nodes.
[Networking] Infrastructure network migration now supports networks with both IPv4 and IPv6 subnets. This significantly reduces the time required to prepare clusters for physical migration and simplifies dual-stack network management.
4. Bug Fixes¶
The cluster join command has a retry issue. (VSTOR-47080)
Added a missing alert for degraded storage arrays. (VSTOR-80691)
Crash reports consume all available disk space on hosts without internet access. (VSTOR-86950)
The check-for-updates task may get stuck. (VSTOR-87198)
The S3 service does not restart after a storage mount restart. (VSTOR-89576)
Router connectivity issues affect multiple routers after a processing error. (VSTOR-91120)
Snapshot deletion causes VM info file synchronization issues. (VSTOR-93041)
NFS storage is created without proper redundancy settings. (VSTOR-94861)
A libvirt race condition causes VM management issues. (VSTOR-95999)
An upgrade fails on older installations that are missing a configuration file. (VSTOR-98221)
Network creation allows spaces in names. (VSTOR-99460)
The disk reservation state persists after an agent restart. (VSTOR-101927)
Error messages are missing from task logs. (VSTOR-102057)
Security fixes. (VSTOR-101281, VSTOR-102955)
DHCP high availability degrades after node replacement. (VSTOR-103562)
A compute node is not removed from the placement service on deletion. (VSTOR-104797)
The Neutron service hangs for five minutes after OVS reconnection. (VSTOR-106311)
The Gnocchi service status is incorrectly reported on secondary management nodes. (VSTOR-106602)
A misleading critical alert appears when the compute backup service is misconfigured. (VSTOR-106690)
The compute backup scheduler allows excessive recovery points (up to 999). (VSTOR-107185)
The node status is incorrectly shown as failed due to a single service failure. (VSTOR-107214)
An upgrade fails when the registry is disrupted during tarball unpacking. (VSTOR-107868)
A race condition exists in network configuration during VM interface attachment. (VSTOR-108710)
A kernel workqueue crash occurs in the network event handler. (VSTOR-108908)
The post-upgrade script accesses unassigned nodes. (VSTOR-109418)
VPN logs are deleted when a router is re-enabled. (VSTOR-109422)
VM unshelving fails after evacuation. (VSTOR-109679)
High availability reconfiguration may get stuck. (VSTOR-109896)
A kernel crash may occur with checksumless chunk servers. (VSTOR-109945)
The disk replacement task blocks system upgrades. (VSTOR-110037)
An upgrade fails due to stuck NFS mount points. (VSTOR-110351)
Manual instance evacuation conflicts with a running evacuation task. (VSTOR-110962)
An incorrect 404 error is returned for pending disk reservations. (VSTOR-110978)
Stability fixes. (VSTOR-113129, VSTOR-115462)
Spice packages are missing in Red Hat Enterprise Linux 9. (VSTOR-113536)
Domain name parsing fails for block nodes. (VSTOR-113808)
Disk auto-replacement is not canceled when the repository is deleted. (VSTOR-113962)
Nested virtualization is misconfigured on AMD processors during upgrade. (VSTOR-114793)
Fixed an S3 recycle bin path validation issue. (VSTOR-115086)
A compute node cannot be added with an OVS VLAN interface without network assignment. (VSTOR-115455)
SR-IOV port attachment does not update the PCI device status. (VSTOR-116308)
A load balancer remains stuck in the deletion state when TLS storage is unavailable. (VSTOR-116344)
A file is not detached on a push error. (VSTOR-116577)
The agent does not use cached CPU information. (VSTOR-116774)
Snapshot creation fails for shelved VMs. (VSTOR-116783)
The Accept-Ranges header is missing from S3 GetObject responses. (VSTOR-117834)
Fixed duplicate quiesce calls during image creation. (VSTOR-117856)
The S3 lifecycle configuration rejects empty filters. (VSTOR-118251)
The backup process blocks storage mount operations. (VSTOR-118451)
A network interface is not removed from the configuration after a detach failure. (VSTOR-118626)
The same node in the HA configuration is repeatedly elected as primary. (VSTOR-118651)
Storage re-encoding is blocked by invalid metadata. (VSTOR-118697)
Failed to delete snapshots of shelved VMs. (VSTOR-118825)
VM tag restrictions prevent underscore and dash characters. (VSTOR-118892)
Fixed an incorrect command for setting password requirements. (VSTOR-119187)
The UI prevents storage policy tier and failure-domain changes. (VSTOR-119333)
Geo-replication creates empty files on unconfigured secondary nodes. (VSTOR-119537)
Compute options are not rolled back on reconfiguration failure. (VSTOR-119570)
OVS Manager creation fails at agent startup. (VSTOR-120216)
A vGPU device is not deallocated during cold migration. (VSTOR-120502)
Router connectivity fails after failover. (VSTOR-120603)