Virtuozzo Hybrid Infrastructure 6.2 (6.2.0-136)¶
Issue date: 2024-06-29
Applies to: Virtuozzo Hybrid Infrastructure 6.2
Virtuozzo Advisory ID: VZA-2024-024
1. Overview¶
In this release, Virtuozzo Hybrid Infrastructure provides a range of new features that cover the compute service, high availability of the management node, object storage management, networking, and monitoring. Additionally, this release delivers stability improvements and addresses issues found in previous releases.
2. New Features¶
[Compute service] Improved Backup and Restore as a Service. Now, with the compute backup service, you have the ability to set quotas for storage space usage and number of backups on external storage systems. Additionally, the newly added statistics for total provisioned and total used storage space for the entire cluster and per project help to improve metering capabilities. Starting from this version, the compute backup service also supports volumes based on external storages such as iSCSI and NFS.
[Compute service] Simplified performance settings for load balancers. During load balancer creation, self-service users have the possibility to select a load balancer flavor and configure access rules based on allowed network CIDR ranges. System administrators can create as many load balancer flavors as required.
[Compute service] Cilium CNI support for Kubernetes version 1.29. Self-service users can use the Cilium network plugin, instead of the default Flannel CNI, for Kubernetes clusters with version 1.29. In this release, the Cilium CNI support is an experimental feature, which is not intended for use in production environments. Starting from the next release, all new Kubernetes versions will be switched to the Cilium CNI by default.
[Compute service] Update of the load balancer image to Ubuntu 22.04 and haproxy 2.4.
[Compute service] In-place upgrade and repair of Windows virtual machines. Added the ability to perform in-place updates of Windows guest operating systems according to the Microsoft recommendations. This can be used to upgrade and repair virtual machines.
[Management node high availability] Primary management node replacement. It is now possible to change the primary management node in the high availability configuration in the user or command-line interface.
[Object storage] Object storage management in the self-service portal. Self-service users have the ability to manage S3 credentials, S3 objects, and properties in the self-service panel, previously used only for compute resource management.
[Networking] IPv6 support for node network interfaces. Cluster networking can be configured to use mixed IPv6 and IPv4 addresses for more efficient IP packet processing and non-limited IP addressing.
[Monitoring] Virtual DHCP connectivity checks. Added a new metric to integrated Prometheus, to show availability of the virtual DHCP servers in the compute cluster.
[Monitoring] Log forwarding with Filebeat. With the built-in Filebeat service, it is now possible to send logs to any external log analysis system.
3. Important Notes¶
Ubuntu 16.04 as a guest operating system will be deprecated in future releases.
4. Bug Fixes¶
Only nodes in the HA configuration should report the HA services. (VSTOR-45749)
Failed to clean up backup slices after promoting the secondary cluster to primary. (VSTOR-56264)
Compute reconfiguration failed after changes in the HA configuration. (VSTOR-61703)
Improved message logging for the MDS Paxos consensus algorithm. (VSTOR-74942)
Added status tracking for the compute service configuration tasks. (VSTOR-77524)
Improved stability of the bulk file delete operation. (VSTOR-78567)
Fixed false-positive alerts on high request error rates during an update. (VSTOR-83747)
When creating a VM in the self-service panel, the default storage policy is not applied. (VSTOR-84265)
Fixed adding new members to load balancer pools that were created bypassing the backend API. (VSTOR-84508)
Some task statuses are not updated in the admin panel after the primary node change. (VSTOR-84752)
Fixed a memory leak in the S3 service. (VSTOR-85081)
Improved message logging and error handling for FUSE. (VSTOR-85086)
Increased the inotify limits for Kubernetes. (VSTOR-85254)
Improved NFS performance by changing the client default configuration. (VSTOR-85296)
A stability fix for PCI devices. (VSTOR-85617)
The chrony package was not installed due to connection issues. (VSTOR-85849)
Fixed pagination for cases with more than 1000 flavors. (VSTOR-85978)
A stability improvement for the Cinder service. (VSTOR-86331)
Errors appear when listing volumes due to incorrect processing of UUIDs. (VSTOR-86365)
Fixed fencing alerts. (VSTOR-86399)
Prevented the storage version from setting to zero. (VSTOR-86429)
Improved checks of spectre mitigation when choosing compute CPU models. (VSTOR-86448)
A stability fix for MDS. (VSTOR-86452)
OVS interfaces are removed from the management node database if they are down. (VSTOR-86491)
Improved protection of user passwords. (VSTOR-86548)
Improved kernel diagnostics via sysrq. (VSTOR-86633)
logrotate is not enabled on some clusters. (VSTOR-86877)
Limited the space consumption by crash reports on hosts without internet access. (VSTOR-86950)
Fixed an incorrect update status in the admin panel. (VSTOR-87330)
Enabled adding one volume to multiple backup plans. (VSTOR-87464)
5. Known Issues¶
During the update, you may receive false-positive alerts. Please wait until the update installation is complete, as the raised alerts will be cleared automatically. (VSTOR-67373)
An inefficient structure of a huge PostgreSQL database may cause the compute API degradation. (VSTOR-86875)