Virtuozzo Hybrid Infrastructure 7.0 (7.0.0-250)¶
Issue date: 2025-06-26
Applies to: Virtuozzo Hybrid Infrastructure 7.0
Virtuozzo Advisory ID: VZA-2025-008
1. Overview¶
In this release, Virtuozzo Hybrid Infrastructure provides a range of new features that cover the compute service, object and core storage, networking, as well as monitoring and alerts. Additionally, this release delivers stability improvements and addresses issues found in previous releases.
2. New Features¶
[Compute service] UI for external storage management. Administrators can now attach, manage, and monitor external storage systems, including Pure Storage®, NetApp®, and any OpenStack-compatible FC/iSCSI/NFS array, directly from the admin panel, reducing CLI usage and minimizing configuration errors.
[Compute service] Default security group templates. Using the OpenStack API, administrators can define project-wide ingress and egress rules that apply automatically upon project creation. This eliminates per-project manual rule setup and minimizes the risk of insecure defaults.
[Compute service] New network port screen. The new screen provides key information about all network ports, helping administrators troubleshoot networking issues in the compute cluster.
[Object storage] IPv6 support. The S3 service now supports IPv6 in addition to IPv4, removing address-space limits in large-scale or dual-stack deployments.
[Object storage] Graphical bucket policy builder. The self-service panel now includes a user-friendly policy builder for creating bucket access policies. Beginners can use guided wizards, while advanced users can paste full JSON policy definitions.
[Object storage] Faster GET throughput. Internal optimizations increase read performance to approximately 12 GB/s per node, significantly accelerating analytics, AI/ML, and media workloads.
[Core storage] Smarter usage metering during erasure coding changes. A tunable overhead parameter prevents temporary usage spikes from causing overbilling or quota violations during large dataset re-encoding.
[Core storage] Automatic SSD/NVMe wear-level detection. New disk health analytics trigger alerts when a disk’s remaining lifetime drops below 10% (warning) or 5% (critical). Once the wearout threshold is reached, affected chunk services are placed into maintenance mode to preserve data redundancy and system availability.
[Core storage] Replication write modes. The storage cluster now supports two write modes for replication redundancy: CS fanout and client fanout. When enabled, these modes optimize storage performance based on network conditions and workload types.
[Networking] Multi-subnet support within a single network. External infrastructure networks can now be assigned multiple IPv4 subnets. This removes the need for additional VLANs and helps prevent public IP exhaustion.
[Monitoring and alerts] Cluster configuration monitoring and alerting. New metrics and alerts enable system administrators to monitor core and management services across cluster nodes and receive notifications about configuration anomalies.
3. Important Notes¶
The Compute API will be unavailable for up to 15 minutes during the update, temporarily disabling the ability to manage compute workloads.
4. Bug Fixes¶
Improved alerts on memory and swap usage. (VSTOR-95964)
A stability fix for block storage. (VSTOR-97082)
Fixed incorrect VPN connection status. (VSTOR-97157)
Stability fixes for the Neutron service. (VSTOR-98157, VSTOR-98509, VSTOR-99169, VSTOR-101941)
Fixed a false-positive alert on port check. (VSTOR-98184)
Fixed the Nginx error “Too many open files.” (VSTOR-99055)
A VIP address can be assigned outside the allocation pool during load balancer creation. (VSTOR-99359)
Fixed a potential infinite loop in the SNMP parsing code. (VSTOR-99624)
A stability fix for Kubernetes. (VSTOR-99870)
Locked archives affect the calculation of space usage. (VSTOR-100409)
Added the ‘tenant_id’ label to volume backup metrics. (VSTOR-100472)
A stability fix for the Gnocchi service. (VSTOR-100484)
Fixed an incorrect self-service URL in the IDP settings. (VSTOR-101119)
Decreased the maximum number of snapshots to 32 per volume. (VSTOR-102028)
Fixed incorrect hypervisor information on compute nodes with similar hostnames. (VSTOR-102694)
Fixed a UEFI boot failure for VMs with multiple virtIO disks. (VSTOR-102711)
A security fix for neighbor VMs in the same network. (VSTOR-103381)
Volumes may be removed from a backup plan when the volume QOS policy is updated. (VSTOR-103429)
A security fix for the Keystone service. (VSTOR-104443)
Fixed EPYC CPU model names. (VSTOR-106482)
A load balancer may get stuck pending updates. (VSTOR-108065)