pocket drop

[shaik]

Hello Proxmox Community,

I am experiencing persistent packet drops on my Proxmox 8.4 server, despite disabling NIC offloads. Here are the details:

**Proxmox Version & Kernel:**
proxmox-ve: 8.4.0 (running kernel: 6.8.12-9-pve)
pve-manager: 8.4.14
Kernel: 6.8.12-9-pve

**Host OS Info:**
Debian GNU/Linux 12

**Network Setup:**
- Host NIC: ens13f3 (1 Gbps, Broadcom NetXtreme BCM5719)
- Connected to Linux bridge: vmbr0
- VMs connected using virtio NICs
- NIC offloads **disabled** (TSO/GSO/GRO off) on host and VMs

**Symptoms:**
- Continuous packet drops under moderate traffic
- Observed while pinging between host and VMs
- Performance issues persist even after disabling offloads

**Hardware:**
lspci | grep -i ethernet
17:00.0 Ethernet controller: Broadcom NetXtreme BCM5719 Gigabit Ethernet PCIe (rev 01)
17:00.1 Ethernet controller: Broadcom NetXtreme BCM5719 Gigabit Ethernet PCIe (rev 01)
17:00.2 Ethernet controller: Broadcom NetXtreme BCM5719 Gigabit Ethernet PCIe (rev 01)
17:00.3 Ethernet controller: Broadcom NetXtreme BCM5719 Gigabit Ethernet PCIe (rev 01)
**Steps already tried:**
1. Disabled TSO, GSO, GRO on host NIC
2. Disabled TSO, GSO, GRO on VM NICs (virtio)
3. Verified bridge configuration

**Request:**
Looking for guidance on:
- Any additional tuning for 1 Gbps Broadcom NICs in Proxmox bridges
- Known issues with this kernel version and Linux bridge
- Debugging steps to identify source of packet drops

 

[weehooey-bh]

Debugging steps to identify source of packet drops

1. Re-review your configuration. Please post your /etc/network/interfaces file.
2. Confirm that the packets are flowing as you expect them to. A packet capture may provide some insights.

Continuous packet drops under moderate traffic

What is your definition of "moderate" traffic?

Performance issues persist even after disabling offloads

Is this the problem you are trying to solve? host-vm network performance? If not, what problem are you trying to solve?