May 15, 2026
Article
Live migration is the process of transferring a running virtual machine (VM) from one host to another with minimal downtime, ensuring continuous access. The brief transition between stopping the VM on the source and starting it on the destination helps avoid service disruption. It is commonly used for maintenance and load balancing, involving the transfer of the VM’s memory, network state, and storage within the cluster.
VMs can be migrated to less utilized hosts when a system is overloaded, improving overall resource utilization and network performance.
The hypervisor abstracts the underlying hardware, allowing VMs to run independently of specific hardware platforms. This enables migration across different systems with minimal or no downtime, though some limitations may apply (e.g., different CPU architectures).
VMs can be consolidated onto fewer hosts during low-demand periods, allowing unused systems to be powered down to reduce energy consumption and costs.
VMs can be relocated to different physical locations to reduce latency or support use cases such as fault tolerance and disaster recovery.
The virtual machine image is stored on an NFS (Network File System) server, as live migration requires shared storage that can be accessed simultaneously by both the source and destination hosts. Both hosts are configured as NFS clients, mounting the same shared storage to ensure the VM image remains consistently accessible throughout the migration process
The VM is booted using the image stored on the NFS shared storage. During live migration, the running VM—including its memory state, CPU context, and device state—is transferred from the source host (Host 1) to the destination host (Host 2) with minimal downtime, while the underlying disk image remains on the shared NFS storage. To maintain seamless network connectivity, a bridge (virbr0) is configured on both hosts, connecting the physical network interface and the VM’s virtual interface. This ensures the VM remains on the same network before and after migration, allowing it to stay reachable without interruption.
The iW-Fibre card acts as the high-speed interconnect between the two hosts, serving as the dedicated network interface through which the VM’s live migration traffic is transmitted. It facilitates the transfer of the VM’s memory pages and state data from Host 1 to Host 2 with low latency and high bandwidth, which is critical for achieving a seamless migration.
The validation was performed using the G35-XFibre-100G platform to showcase SmartNIC-enabled live migration capabilities, highlighting seamless workload mobility and high-performance virtualization.The solution is built on the OpenNIC design, with the OpenNIC kernel driver enabling migration through the SmartNIC datapath. Virtual machines were instantiated and managed using QEMU and Virt-Manager, ensuring a flexible and robust virtualization environment.
A centralized storage environment was established by creating and exporting a shared directory via NFS, followed by server initialization and validation. The VM image was hosted on the shared storage to enable cross-host accessibility.
Virtual machines were initiated across source and destination hosts, with the source actively running the VM and the destination pre-configured with an identical environment to ensure migration readiness
The virtual machine was seamlessly transitioned from the source to the destination host with minimal downtime, demonstrating high system resilience and uninterrupted operational continuity.
In conclusion, live migration ensures minimal downtime while moving VMs between hosts, enabling continuous service along with efficient maintenance and load balancing.
For platform evaluation or additional information, contact us at mktg@iwave-global.com
We appreciate you contacting iWave.
Our representative will get in touch with you soon!