Product release: Virtuozzo Hybrid Infrastructure 4.0 (4.0.0-734)

Issue date: 2020-10-06

Applies to: Virtuozzo Hybrid Infrastructure 4.0

Virtuozzo Advisory ID: VZA-2020-063

1. Overview

In this release, Virtuozzo Hybrid Infrastructure provides a wide range of new features that enhance the end-user experience and service providers’ interoperability. The improvements cover networking, storage core, appliance, object storage, monitoring, performance charts, and localization. Additionally, this release delivers stability improvements and addresses the issues found in previous releases.

2. New Features

  • [Compute] Rescue mode for virtual machines. Rescue mode provides a mechanism to access the virtual machine’s boot volume in case the virtual machine experiences boot problems.

  • [Compute] Distributed virtual router. The virtual routing architecture has been changed from centric to distributed (across all compute nodes).

  • [Compute] QoS policies for virtual machine volumes. The storage policy has been extended with additional parameters that define the IOPS and bandwidth limits of virtual machine volumes.

  • [Compute] Paging for compute and common objects in the admin panel. Improved the admin panel performance for large numbers of virtual machines.

  • [Compute] VLAN-based network management for virtual machines. Simplified the process of creating VLAN-based networks for virtual machines.

  • [Compute] More supported guest operating systems. Added support for Debian 10, CentOS 8, and Ubuntu 20.04.x.

  • [Integration] CloudBlue Connect. Service providers can now integrate with the CloudBlue Connect billing and provisioning system. Integration supports both the flat and pay-as-you-Go billing models.

  • [Networking] Exclusive traffic type reassignment and cluster IP address change. It is now possible to change the configuration of exclusive traffic types or IP addresses in the cluster. This is needed when the node IP address has changed but network topology remains the same, or if the network subnet has changed. It is also possible now to reassign a service traffic type when a customer wants to relocate a service to another network (for example, move the “Storage” traffic type to a dedicated network).

  • [Core storage] Node locations as failure domains. A failure domain defines a scope (a location) within which services may go down together. Now nodes can be assigned to specific locations: rooms, rows, and racks. Locations provide redundancy among failure domains. This gives an additional, flexible way to service providers to utilize the available hardware more efficiently while delivering high availability.

  • [Core storage] Improved latency and performance. Kernel and networking optimizations.

    • Erasure coding is now available for running virtual machines.

    • Erasure coding self-healing now has less impact on performance.

    • The storage networking subsystem has been improved for lower I/O latencies and CPU consumption.

    • RDMA performance and stability have been improved. RDMA is now a production-ready feature.

  • [Object storage] S3 geo-replication improvements.

    • Improved security for geo-replication (bucket ACLs).

    • User-level and bucket-level customization of S3 geo-replication.

  • [Object storage] S3 bucket access log. Now the records of requests to a bucket can be provided and stored in another bucket. In addition, end users can now enable the access log via Amazon S3 API and get detailed information about the objects in their buckets.

  • [Monitoring and alerting] A dashboard for S3 performance and statistics charts. You can now monitor:

    • Services, including S3 GW, NS, OS services availability, rate of operational requests, bandwidth, as well as GET and PUT latencies.

    • Geo-replication, including service availability, error rates, replication backlog, queue depth, and bandwidth.

  • [Monitoring and alerting] Dashboard for performance and statistics charts for NFS. You can now monitor the availability of NFS servers and services, usage, shares, exports, latency, IOPS, and bandwidth.

  • [Localization] Chinese localization for the admin panel and documentation.

  • [Localization] Japanese localization for the admin panel and documentation.

  • [Other enhancements] Disabled pNFS. pNFS is now disabled by default and cannot be enabled for any forthcoming use cases.

3. Bug Fixes

  • [Compute networking] After virtual router creation, the internal IP address goes down. (VSTOR-33859)

  • [Compute service] The docker service has no log rotation. (VSTOR-32173)

  • [Compute service] Floating IPs and virtual routers are not available after removing nodes from the compute cluster. (VSTOR-30506)

  • [Compute service] The dashboard service may get stuck in a restart loop. (VSTOR-32120)

  • [Compute service] Storage space usage in a project is calculated incorrectly for volumes with snapshots after their storage policy has been changed. (VSTOR-34054)

  • [Compute service] Kubernetes cluster creation hangs if it cannot access the public Docker hub. (VSTOR-32238)

  • [Compute service] The noVNC console sends duplicate keystrokes to Windows virtual machines. (VSTOR-30975)

  • [Compute service] Unable to remove a load balancer while in the “deployment_failed” state. (VSTOR-35034)

  • [Compute service] Cannot deploy the compute cluster after reassigning the “Internal management” traffic type. (VSTOR-33202)

  • [Compute service] Load balancers are in the “error” state after the update to version 3.5.1. (VSTOR-32237)

  • [Core storage] The archive server client crashes during reconnection. (VSTOR-33017)

  • [Core storage] Links between a frontend chunk object and an internal journal file system object may be broken. (VSTOR-33577)

  • [Core storage] Avoid unnecessary reconfiguration of allocated disks. (VSTOR-31951)

  • [Core storage] Storage services deployed on the same disk generate an exception in logs. (VSTOR-32782)

  • [Core storage] The same disk can be assigned twice in two parallel bulk assign tasks. (VSTOR-33012)

  • [Core storage] After assigning roles to a node disk during the cluster creation and reassigning them afterwards, the node may appear as “Failed.” (VSTOR-30931)

  • [Admin panel high availability] An unnoticed failure to replicate the management node database may cause the compute metadata loss. (VSTOR-34723)

  • [Backup storage] The Backup Gateway geo-replication command breaks compatibility. (VSTOR-34891)

  • [Backup storage] The Backup Gateway utility does not correctly accept a password containing spaces from stdin. (VSTOR-32028)

  • [Object storage] Unable to start replication of a bucket with a large number of objects. (VSTOR-27097)

  • [Object storage] It is possible to remove the single node with the S3 configuration from the S3 cluster. (VSTOR-30376)

  • [Object storage] The NFS remains active and running after the node has been released from an NFS cluster. (VSTOR-30622)

  • [Updates] The upgrade from version 2.5 to 3.0 fails due to the installed GeoIP module. (VSTOR-31734)

  • [Updates] Failed to download updates due to a newer version on the update server. (VSTOR-35014)

  • [Updates] Under certain circumstances, downloading packages may fail. (VSTOR-33446)

  • [Updates] After an update with enabled high availability, the admin panel may become unavailable. (VSTOR-26542)

  • [Installer] During installation, the specified hostnames are converted to lowercase. (VSTOR-35352)

  • [User interface] The problem report location is missing if there is no Internet connection. (VSTOR-34770)

  • [User interface] Improve the text for the “Incorrect journaling settings” alert. (VSTOR-34387)

  • [User interface] Email notifications sent via SMTP do not contain the cluster name. (VSTOR-32933)

  • [User interface] Unable to interpret special symbols in passwords. (VSTOR-30216)

  • [User interface] Several services are not configured to restart automatically. (VSTOR-20993)

  • [User interface] The old product version is displayed after a successful update. (VSTOR-25662)

  • [User interface] License capacity is shown as ”Total” in the physical space widget. (VSTOR-27237)

  • [User interface] A 25 Gbit network card is reported as a 10 Gbit network card in the admin panel. (VSTOR-9422)

  • [User interface] Redis stores superfluous failed task metadata. (VSTOR-36654)

  • [User interface] NFS4 is not selected by default in the admin panel. (VSTOR-33809)

  • [User interface] Network counters differ in the admin panel and inside a load balancer virtual machine. (VSTOR-32756)

  • [User interface] There is no “Security” field validation when saving email settings. (VSTOR-14465)

  • [User interface] Check boxes are too bright on very light displays. (VSTOR-6694)

4. Known Issues

  • [Core storage] The chunk service fails during a journal resize. (VSTOR-30111)

  • [Core storage] Global Catalog in Acronis Cyber Backup does not reclaim the last chunks of RAID6 files. (VSTOR-34465)

  • [User interface] The management node does not return a clear backup status if high availability is enabled. (VSTOR-32254)

  • [User interface] The admin panel does not update the node status while the backend is busy. (VSTOR-35443)

  • [User interface] Placements associated with images are displayed in the virtual machine creation wizard. (VSTOR-35900)

  • [Installer] In the installation wizard, it is not possible to turn on the network time if it was turned off on the previous step. (VSTOR-30581)

5. Installing the Update

You can upgrade Virtuozzo Hybrid Infrastructure 3.5 to 4.0 in the SETTINGS > UPDATE section. A reboot is required to complete the upgrade. Upgraded nodes will be rebooted automatically, one at a time. During the reboot, the storage service and the admin panel may be unavailable on cluster configurations without the redundancy of services or data.

IMPORTANT: During the upgrade, custom public endpoints for the OpenStack API that use domain names instead of IP addresses will be replaced with the default values. This can lead to OpenStack API unavailability if these endpoints are used to connect third-party software or to manage the infrastructure. To make the endpoint configuration persistent after the upgrade, follow the instructions in Setting a DNS name for the compute API.

The JSON file with the list of new and updated packages is available at