Hyper Converged Infrastructure (HCI/aSV)

Sangfor HCI and aSV provide a unified infrastructure combining compute, storage, networking, and built-in security to simplify deployment, operations, and services.
{{ $t('productDocDetail.guideClickSwitch') }}
{{ $t('productDocDetail.know') }}
{{ $t('productDocDetail.dontRemind') }}
6.11.1R1
{{sendMatomoQuery("Hyper Converged Infrastructure (HCI/aSV)","Enable Turbo Service")}}

Enable Turbo Service

{{ $t('productDocDetail.updateTime') }}: 2026-01-05

Description

HCI carries many customers’ critical business services, including database applications, big data analytics, artificial intelligence, and machine learning. These services require high storage IO performance, which is directly associated with service processing speed and user experience. For this reason, we integrated HCI with the SPDK solution, significantly improving the performance of VMs running IO-intensive services.

The Storage Performance Development Kit (SPDK) is a toolkit for building high-performance storage applications. HCI uses the SPDK vhost at the virtualization layer to enable VMs to run in user mode so that IO streams can bypass the kernel to directly access the hardware, reducing the resources consumed by scheduling. The Turbo service based on SPDK vhost reduces the average latency of reading and writing large blocks of VMs from 4 ms to 2 ms, improves throughput efficiency by about 50%, and IOPS of small blocks by more than 20%, resulting in a significant increase in performance.

We deeply optimized the process of SPDK for accessing virtual storage, enabling IO at a single depth to directly access virtual storage in the context of round-robin threads, which reduces thread switching overhead, and switching IO at different depths to pipe threads to divide large blocks into multiple IO requests of 128 KB to concurrently process them. The IO performance test shows that compared to the original solution, the VM performance has been improved by 20% to 50%. We also transformed the SPDK framework to make it adapt to advanced features such as double instance checks, HA, and in-place active upgrades for VMs, ensuring a smooth SPDK VM experience.

Ordinary VMs and VMs using SPDK can coexist on HCI and convert to each other. You can choose whether to use SPDK based on the VM services, enabling a balance of performance and functionality.

IMG_256

Precautions

SPDK communicates with SPDK VMs through VirtIO protocol. To create an SPDK VM on a Windows system, create an ordinary VM and install vmTools on it first, then convert it to an SPDK VM. Linux system comes with the VirtIO driver, you can directly create an SPDK VM on it.

It is not supported to create backups (including agentless backups) and CDP policy for SPDK VMs on HCI. To back up SPDK VMs, use a third-party agent or back up through the application layer.

It is not supported to perform disaster recovery for SPDK VMs on HCI and SCP. To recover SPDK VMs, use a third-party plan or recover through the application layer.

The huge-page memory feature must be enabled and the memory reclaiming feature must be disabled for SPDK VMs; otherwise, the full memory page may not be found when memory is overcommitted, causing VMs to fail to start up.

SPDK VMs can be migrated to another node associated with the same datastore when running and to another datastore when powered off.

SPDK VMs cannot run across datastores.

SDPK VMs do not support the encryption features.

SPDK VMs can only be stored on virtual storage, not on external or local storage.

SPDK VMs can be created through the full deployment method, not from templates.

Turbo mode cannot be enabled for VMs with snapshots.

SPDK VMs cannot use physical disks and virtual shared disks.

SPDK VMs use RAW disks. There is no need to change the disk allocation type when the VMs are running because there is no performance difference between thin provisioning and pre-allocating. Dynamic provisioning is not supported by SPDK VMs.

SPDK only supports datastores using three or more nodes.

Read caching cannot be enabled for SPDK VM disks.

Space reclamation cannot be enabled for SPDK VM disks.

Disk based snapshots cannot be created for SPDK VMs.

SPDK VMs can only be cloned across datastores when powered off, and the destination datastores must use at least three nodes.

SPDK VMs cannot be suspended.

SPDK VMs cannot be imported or exported.

Steps

Go to Systems > Advanced > Virtual Machine, select Enable of Turbo Service, and click Save.

Create a VM by referring to Section 5.1.2 Create A New Virtual Machine.

Click Power Off of the created VM, then click its More, go to Edit > Configuration > Memory, select Enable huge-page memory, and then click OK. Please make sure that the VM has vmTools installed.

Click More of the VM again, and go to More > Enable Turbo Mode.

Confirm the information on the pop-up dialog, select I have understood the risks, and click OK.

Go to Tasks to view the conversion progress.