SvHCI FAQ
The StorMagic SvHCI FAQ attempts to answer a wide variety of queries relating to StorMagic SvHCI, its deployment and features.
In addition to this FAQ, complete documentation for the deployment of SvHCI is available to read by visiting the SvHCI manual page.
Have you read through the SvHCI data sheet, and the SvSAN Technical Overview white paper, both of which are packed with useful information about SvHCI, its requirements, and capabilities?
If you cannot find an appropriate answer within the SvHCI FAQ, you can contact the StorMagic team at [email protected].
Question Categories
SvHCI Architecture
SvHCI is a full-stack HCI solution since it includes a hypervisor, virtual networking, virtual storage, and management. It is a complete virtualization platform for x86-based physical servers.
SvHCI is a Linux-based appliance, based on KVM as the hypervisor, Open vSwitch for virtual networking and StorMagic SvSAN™ for storage.
The complete index of compatible hardware for SvHCI can be found in the Hardware Compatibility List (HCL): https://stormagic.com/doc/svhci/2-4-0/en/Introduction/hcl.htm
We are working hard to test as many combinations and permutations of hardware, controllers and drivers, and if there are specific components or drivers that you would like to see supported, please contact your StorMagic representative who will be able to pass on your feedback and add it to our testing and validation effort.
SvHCI supports up to 50 virtual machines (VMs) per cluster if all VMs require high availability (HA). If HA is not required, then each node can support up to 50 non-HA VMs. Therefore, a 2-node cluster would permit up to 100 non-HA VMs.
For environments with a mix of HA and non-HA VMs you can use this formula:
A = # of VMs that require high availability
B = # of VMs that do not require high availability
2*A + B ≤ 100
SvHCI supports the following list of guest operating systems: https://stormagic.com/doc/svhci/2-4-0/en/Introduction/guest-os.htm
High Availability
A total of 2 nodes are required to create highly available SvHCI clusters.
StorMagic does not guarantee a level of availability provided by our software. However, we design our software for 100% uptime and have many examples of customers going years without a single instance of downtime.
SvHCI’s mirroring synchronizes all of the data between the two nodes so that it is always on both servers. So if a server fails, its VM data as well as the datastores are already on the other server and therefore there is no wait time to get that data available and online.
Regarding the restart of VMs on the surviving node, there is a small delay of around 30 seconds. This compares favorably to VMware and Microsoft hypervisors which take around 2.5 minutes and 4 minutes respectively to restart VMs on another node in the event of a failure.
SvHCI is an active-active 2-node cluster. During configuration, VMs can be selected to run on node A, or node B, to create a balanced system and in the event of a node going offline, all the VMs from the failed node will migrate over to the surviving node.
Yes, SvHCI allows the two servers in a cluster to be separated physically. This could be in alternate racks, separate rooms in a single building, or separated over greater distances over a campus or metro area. This is known as a stretched or metro cluster, of which more information is available here.
No. Live migration of VMs works with all servers on the SvHCI Hardware Compatibility List (HCL) which is available here.
Please see the Technical Product Documentation section for this topic here.
Storage
SvHCI supports any physical hardware RAID controller for servers on the SvHCI HCL and therefore supports any RAID configuration through that controller. If there is no physical hardware RAID controller available, SvHCI offers software-enabled RAID 0, 1 and 10. In addition, SvHCI supports Intel Virtual RAID on CPU (VROC) embedded storage RAID controllers for RAID 0, 1 and 10.
No, SvHCI does not support connections to external storage at present, though it is on the product development roadmap. However, the reverse is possible: SvHCI can present an iSCSI target that can be used by other compute nodes.
Networking
The minimum network connection to an SvHCI node is a single 1GbE port. However, the higher the network bandwidth between SvHCI nodes, the better. SvHCI supports up to 100GbE network interface cards and ethernet speeds. For full system requirements, refer to the SvHCI Data Sheet here.
Yes, SvHCI supports back-to-back network cabling.
Management
Yes, SvHCI is managed through a web-based graphical user interface known as the SvHCI Virtualization Manager. Furthermore, StorMagic’s centralized fleet management product, Edge Control, allows management of an organization’s entire fleet of SvHCI nodes and clusters from a single web console.
The SvHCI Virtualization Manager is included and installed on each server running SvHCI. To access the user interface, simply open a web browser and enter the IP address for the specific SvHCI host.
StorMagic’s centralized fleet management product, Edge Control, allows management of an organization’s entire fleet of SvHCI nodes and clusters from a single web console.
The SvHCI Virtualization Manager management console is part of the SvHCI software stack running on each server. The server would continue running if the Virtualization Manager is not accessible. The Virtualization Manager process can fail, but it is separate from the core server virtualization service which may continue running. However, in a single node scenario, if the entire server went down this would mean that the Virtualization Manager, virtualization service and VMs would fail.
No, not in the current version of SvHCI. Security enhancements are being prioritized as part of the product roadmap however.
No, this is not supported in SvHCI at present. It is an item being considered for the development roadmap.
Yes, through a deployment automation workflow, SvHCI software can be installed remotely onto bare-metal, headless servers allowing rapid, repeatable rollout across large site counts.
The workflow includes staged rollout support (pre-stage in lab, complete onsite), headless execution with lightweight logging, and optional auto-shutdown on completion. Status, success, and error signals can be sent to a configurable endpoint to help track installation progress across many hosts. Once installation completes, bespoke post-install configuration scripts can be run; SvHCI 2.4 provides API support for common configuration activities along with sample scripts.
Backup & Data Protection
SvHCI supports agent-based backup with any third-party backup software product. This is achieved by installing a backup agent in any VM that must be backed up. Storing the backup in another storage location is the responsibility of the user. In the event of disaster recovery, the user can recover the VM from backup.
No, SvHCI does not support agentless backup at this time, but it is being actively worked on for future release. In the interim, we highly encourage users to have a full understanding of the pros and cons of the agent-based approach to backup. Read this blog post for more information: https://stormagic.com/company/blog/agent-vs-agentless-backup/
Yes, SvHCI supports VM-level snapshots for crash-consistent recovery, with a maximum of 16 virtual machine (VM) snapshots retained for each VM. You can use snapshots to create a point-in-time (PIT) copy of data and capture the operating system, software application, or disk state. VM disk data and virtual machine configuration settings are retained, enabling rollbacks to a previous point-in-time for quick recovery.
Performance
SvHCI currently commits the memory allocated to a VM. This is the safest option, but we understand that competitors can overcommit memory, and therefore it is a development that we are considering as part of our product roadmap.
No, there are no restrictions. SvHCI allows you to run any application inside the guest OS running on a VM. However, we recommend sizing the correct system environment and performance to your needs.
Pricing & Licensing
SvHCI is licensed by the node (virtualized server) and the amount of storage needed. The SvHCI license is available in storage tiers of 2TB, 6TB, 12TB, 24TB, 48TB, and unlimited TB options.
Yes. SvHCI pricing includes the hypervisor and support for the entire software stack, all as a subscription.
Migrating to SvHCI
Yes, VMs can be imported into SvHCI.
For VMware environments, the VMware VM Import Utility within SvHCI provides a scripted process to bring VMs from VMware into SvHCI, with VM configuration settings and all virtual disks with data copied over. VMs can be transferred using warm import, allowing most VM data to be transferred while the source virtual machine remains live in the VMware vSphere environment. Furthermore, imports can be queued and run sequentially.
For other virtualization platforms, users have two options. They can import VMs in the standard OVF/OVA file format using SvHCI’s VM Import/Export feature, or they can convert the VMs into a raw disk file and then import them via the “Upload disk image” feature under “Virtual Machines”. For environments with a large number of clusters, StorMagic Technical Services can assist with scripting the VM migration for ease.
No, with the current release of SvHCI, the import of VMs must be performed one by one.
Yes, VMs can be deployed into SvHCI in OVF and OVA formats, using the VM Import/Export feature.
Roadmap
No. Like most technology vendors, the SvHCI product roadmap is not publicly available. It is available under NDA and shared only on a case-by-case basis with customers, prospects, channel or alliance partners. Please contact your StorMagic sales representative to arrange a discussion about the SvHCI roadmap.
Yes, StorMagic operates an ongoing Beta Program for SvHCI that allows new or existing StorMagic customers, channel partners and OEMs to evaluate upcoming product features and provide feedback to the development team. Program participants receive access to time-limited 30-day SvHCI beta software licenses that are fully-featured with unlimited TB storage capacity. The beta licenses must only be used for pre-production, lab deployments with test workloads. Deploying the software in production environments and/or running production workloads on the software will not be supported by StorMagic. Please contact your StorMagic representative or use this form to request more information about the Beta Program.