Virtuozzo Hybrid Infrastructure installed on bare-metal servers unites them into a single cluster, which can be easily scaled by adding disks or nodes. The cluster is managed via a highly available web-based admin panel and the command line. The admin panel provides extensive monitoring of all components. The overview dashboards are integrated with Prometheus, Grafana, SNMP, and Zabbix, to provide an insight into the infrastructure status. In addition, the alerts system keeps the administrator informed about misconfiguration, failures, and other issues.
Clustering helps avoid data loss with replication and erasure coding. With high availability enabled, the cluster and services have no single point of failure. The storage cluster is self-healing: if a node or disk fails, the cluster will automatically try to restore the lost data. Besides, with non-disruptive rolling updates the data stays available even when updating the nodes. In case of a node maintenance or applying hotfixes, the workload is migrated to other available nodes.
This section describes the major infrastructure components and their architecture: the storage and compute clusters, as well as backup, block, object, and file storage.