Virtuozzo Hybrid Infrastructure 6.1 (6.1.0-238)¶
Issue date: 2024-03-28
Applies to: Virtuozzo Hybrid Infrastructure 6.1
Virtuozzo Advisory ID: VZA-2024-013
1. Overview¶
In this release, Virtuozzo Hybrid Infrastructure introduces a new service—Backup and Restore as a Service—as well as provides a range of new features that cover improvements in the compute services and object storage. Additionally, this release delivers stability and security improvements, and addresses issues found in previous releases.
2. New Features¶
[Compute service] Integrated Backup and Restore as a Service. The new backup service allows you to back up VM volumes, and then restore virtual machines and volumes from them. This helps to prevent data loss and mitigate issues with software or hardware. The service supports external and integrated storage (NFS and Object Storage) as a backup destination, thus eliminating the need for purchasing an external backup solution and decreasing the product TCO.
[Compute service] New features for load balancers. Added support for the HTTP headers, PROXY protocol, TLS, and certificate upload for the load balancing service.
[Compute service] External iSCSI storage support. Starting from this release, the compute service supports any external iSCSI storage system that has the OpenStack Cinder driver. You can find the list of available iSCSI drivers at https://docs.openstack.org/cinder/xena/reference/support-matrix.html.
[Compute service] Support for new Kubernetes versions: 1.26, 1.27, 1.28. Starting from this release, the Kubernetes release support policy will be the following:
We support only the four latest minor versions. For example: 1.28, 1.27, 1.26, and 1.25.
We support only the latest patch for the supported minor version.
We provide support for a new minor version during a six-month period after the upstream release.
[Compute service] Baseline CPU model per node. To combine multiple generations of the same CPU vendor, you need to set the baseline CPU model based on the oldest generation, to make all CPUs compatible with each other. Previously available only for the entire compute cluster, the baseline CPU model now can be set on the per-node basis. Thus, you can have multiple sets of CPU generations in a single cluster covered by multiple CPU models.
[Compute service] Improved domain quota management. Now, you can set domain quotas for vCPUs, RAM, and storage policy space in the admin panel.
[Compute service] Improved monitoring for virtual machines. The new monitoring charts available in the admin and self-service panels show separate metrics for each VM volume and network interface.
[Compute service] SSL protocol support for Kafka connection. Added the possibility to configure the SSL protocol for Kafka connection, to secure message transfer between your cluster and metering solution. SSL can be configured for encryption and also serves as two-way authentication between the Kafka server and client.
[Compute service] New guest OS support. Added support for SUSE Linux Enterprise 15 (SP3, SP4, SP5) as a guest operating system.
[Object storage] Performance improvements. Added the possibility to place NS and OS journals on a faster storage tier than is used for data, to improve S3 storage performance.
3. Bug Fixes¶
Introduced partially synchronous writes of memory cache limits to catch potential errors at an early stage. (VSTOR-59718)
Creating a snapshot may get blocked by the process of deleting another snapshot. (VSTOR-61552)
Increased the speed of volume resize operations. (VSTOR-61692)
Fixed an issue with the IPsec bypass rules. (VSTOR-66435)
Add recurrent tasks for handling snapshots that were not deleted completely. (VSTOR-74476)
A snapshot is created incorrectly if the previous top snapshot was not removed completely. (VSTOR-74824)
Improvements in the email notification configuration. (VSTOR-75647, VSTOR-79808)
After a failed attempt to release a node from the backup storage, the retry is not requested until the node is back online. (VSTOR-75699)
A cluster node crashed due to a race condition. (VSTOR-76384)
Created a log record for a load balancer moving into the pending state. (VSTOR-76628)
Live migration failed after aborting the blockcommit job. (VSTOR-76813)
The chunk service in the direct_io mode fails on drives with 4k sectors. (VSTOR-77130)
The admin panel shows the IP address for a domain if the DNS name is configured for the compute API. (VSTOR-77131)
A storage policy name cannot be changed in the admin panel if it has a space at the end. (VSTOR-78315)
NFS container archive validation may fail. (VSTOR-78479)
A stability fix for the chunk service journal. (VSTOR-78664)
Degradation of write speed to the backup storage with the Amazon S3 destination on high-speed connections. (VSTOR-79332)
A virtual machine may get stuck in the shelving status. (VSTOR-79378)
Fixed the filter value in the Neutron service log. (VSTOR-79494)
A network interface may go offline after changes in its parameters. (VSTOR-79909)
An inactive event loop in the S3 account control service. (VSTOR-79981)
The S3 gateway crashes when trying to generate too many access keys. (VSTOR-80092)
A stability improvement for the kernel updates. (VSTOR-80323)
Irrelevant nodes are mentioned in the S3 service alerts. (VSTOR-80743)
Improved the size calculation of erasure coding files. (VSTOR-81226)
Some geo-replication metrics may drop below zero and cause an S3 service failure. (VSTOR-81326)
Changing properties of an interface with VLANs results in the loss of dependent routes. (VSTOR-81343)
Fixed an issue with incorrect trap names in the Zabbix template. (VSTOR-81530)
Improved parsing options for commas in vinfra. (VSTOR-82493)
Listing domain users fails if an admin user has locked IP addresses. (VSTOR-82506)
Cluster total space increases after the chunk service restart. (VSTOR-82639)
Fixed an incorrect subnet value in the documentation. (VSTOR-82983)
iSCSI volumes may stop responding if moving to the Standby mode was not completed before setting the Active path. (VSTOR-83102)
4. Known Issues¶
A stability issue with RDMA under a high load. If you use RDMA, please wait for the next update in order to update the cluster without any issues. (VSTOR-83964)
Domain storage policy quotas are not validated during the creation of a Kubernetes cluster. (VSTOR-81220)
The cluster is not available via the virtual IP address after a failover during the HA reconfiguration. (VSTOR-82626)
The “Nodes” tab is not displayed properly after creating target groups with several volumes. (VSTOR-83316)
Cannot select multiple times per day in the backup plan schedule. (VSTOR-83416)
Failed to retry upgrading a single-master Kubernetes cluster with insufficient quotas for temporary master nodes. (VSTOR-83678)
The “Backend task state” dashboard does not show when the task was completed. (VSTOR-83687)