Create a virtual data center with software-defined infrastructure
vStack HCI architecture
The vStack HCP technology stack
VStack SDN has three ways to deliver virtual networks:
- VLAN (VLAN);
- VXLAN (VLAN);
- GENEVE (RFC 8926: Tunnel protocol, native implementation)
Each particular virtual network instance is based on a distributed switch, has its own MTU and supports jumbo frames, TSO/GSO, TCP MSS clamping and path mtu discovery.
- More control with more flexibility
- Customizable infrastructure and virtual resource allocation for real-time changes to the network infrastructure
- Devices with multiple levels of security can be isolated and zoned to separate devices
- High performance: at least 22 Mps
- Heterogeneous network support out of the box
The virtual machine consists of vCPU, RAM, virtual network ports and disk devices. The lightweight nature of vStack CPU Overhead or the performance degradation of a virtualized server compared to a physical server is minimal. The vStack SDC performs efficiently in CPU overcommit environments and offers better performance than its peers. Cost efficiency of up to 900 percent in the SDC layer.
Comfortable platform operation in the vStack SDC layer is supported:
- autonomous and self-sufficient CPU overcommit;
- real-time vCPU resource limiting;
- ability to create virtual machines with flexible customization;
- Customizability of guest OS in cloud images;
- Limiting real-time network and disk performance;
- Capability to create snapshots of VM configuration.
vStack SDS provides data storage with redundancy and transactional integrity and adds automatic data consistency recovery features. The layer consists of pools that include the disks of each node. At any given time, the pool is running on a particular node in the cluster. One of the key attributes of a cluster is redundancy, where more elements than required are used in the operation of the infrastructure. This allows for redundancy. If one node fails, it will be excluded from the cluster. In this case, the pools will lose one disk each. After that, thanks to the failover procedure, the pool from the failed node and all virtual machines running on it will be available on the other node. The failover is an automatic process.
Tasks performed by SDS:
- creates basic storage entities (pools) with cluster redundancy;
- provides storage primitives (virtual machine disks) to the SDC layer;
- compression and reduplication;
- maintains internal data integrity;
- creates clones and snapshots; self-heals data;
- maintains transactional integrity.
vStack Management is a service virtual machine that contains software for consistent storage and modification of system state, component performance information, and provisioning of endpoint APIs. The primary purpose of vStack Management is infrastructure management.
In addition to system state, the purpose software stores performance information about layers such as virtual machines.The goals of vStack Management are:
- Consistent storage of the infrastructure object layer.
- Holds component performance information.
- Manages the entire infrastructure and its individual components via a web interface and API.
vStack OS is a set of interconnected programs that are used to manage the hardware resources of the vStack hyperconverged platform and also to provide standardized primitives (a set of instructions to perform specific functions) that are used by applications.
In other words, vStack OS is a layer needed to enable software and hardware usage. The fundamental properties of vStack OS are similar to those of other operating systems.
Do you still have questions?
vStack HCP application scenarios
Register as a partner
Please fill out the registration form. We will contact you shortly to arrange the type of partnership that works best for you.