I’ve troubleshot this to the best of my ability and I’m uncertain I have the knowledge to resolve the issue. Here’s the problem:
Large file transfers (> 1GB) between Windows 10/11 VMs and my TrueNAS storage pool will stall, resume, stall, resume, etc. until the file transfer completes.
My Setup:
- System 1 (NAS) (S1): TrueNAS Dragonfish, several VMs with both Linux and Windows VMs, storage pool is a (6) 14TB in raid-z2 configuration, (2) NICs with all VMs using the NIC 2 (with TN using NIC 1)
- System 2 (Proxmox) (S2): Several VMs with both Linux and Windows VMs, (2) NICs with all systems using bridges
What I know:
- The issue does not exist for file transfers between bare metal Windows 10/11 installations and the pool. It only occurs when transferring files between Windows 10/11 VMs only.
- The issue does not exist for file transfers between Windows SERVER VMs and the pool. It only occurs when transferring files between Windows 10/11 VMs and the pool.
- Where the Windows VM sits is irrelevant. Whether on system 1 or system 2, a Windows 10/11 VM transferring a large file (using SMB) from S1 storage pool will exhibit this behavior.
- Direct transfers in Linux (using scp command) do not show this behavior. They are smooth and cap at the network transfer limit.
- Direct transfers in Windows Server (2016) VMs do not show this behavior. They are smooth and cap at the network transfer limit.
- All Windows VM NICs are setup using virtio red hat ethernet driver.
The only issue is with virtual machines running Windows 10 or 11 installations.
Does anyone have an idea what might be the issue?