Hi everyone!
I have:
Proxmox 9.1,
sfp ens13f0np0 Speed: 10000Mb/s, MTU 9000, physical NIC,
Bridge vmbr0 Speed: 10000Mb/s, MTU 9000
Single VM with Windows 2025 and Red Hat VirtIO Ethernet Adapter 10 Gbps. (VirtIO (paravirtualized)). All last Virtio drivers for Win was installed.
Problem: large dropped packets on vmbr0 and win nic. Its critical for damn MS Exchange (ya, i known). Statistics:
ens13f0np0:
vmbr0:
windows nic:
As the statistics show, on Windows, 2 million out of 20 million packets were lost, or one in ten.
Uptime: 3 days.
On Proxmox:
Uptime: 19 days, 63 million packets were sent to vmbr0, 680,000 were lost.
On the physical ens13f0np0: 233 million were sent, 85,000 were lost.
It seems like packets from the Windows VM aren't reaching the physical interface and are being dropped somewhere.
I have:
Proxmox 9.1,
sfp ens13f0np0 Speed: 10000Mb/s, MTU 9000, physical NIC,
Bridge vmbr0 Speed: 10000Mb/s, MTU 9000
Single VM with Windows 2025 and Red Hat VirtIO Ethernet Adapter 10 Gbps. (VirtIO (paravirtualized)). All last Virtio drivers for Win was installed.
Problem: large dropped packets on vmbr0 and win nic. Its critical for damn MS Exchange (ya, i known). Statistics:
ens13f0np0:
Code:
ens13f0np0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc mq master vmbr0 state UP mode DEFAULT group default qlen 1000
link/ether 6c:83:75:0e:5d:f4 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
220619603253 233473964 0 85390 2605 11798108
TX: bytes packets errors dropped carrier collsns
89745549722 97421350 0 0 0 0
vmbr0:
Code:
vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc noqueue state UP mode DEFAULT group default qlen 1000
link/ether 6c:83:75:0e:5d:f4 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
17500241620 63673589 0 680897 0 11660146
TX: bytes packets errors dropped carrier collsns
682621269 2341953 0 0 0 0
windows nic:
Code:
IPv4 Statistics
Packets Received = 20762904
Received Header Errors = 0
Received Address Errors = 277016
Datagrams Forwarded = 0
Unknown Protocols Received = 7
Received Packets Discarded = 2361699
Received Packets Delivered = 18124984
Output Requests = 14285872
Routing Discards = 0
Discarded Output Packets = 7
Output Packet No Route = 0
Reassembly Required = 1054
Reassembly Successful = 286
Reassembly Failures = 0
Datagrams Successfully Fragmented = 0
Datagrams Failing Fragmentation = 0
Fragments Created = 0
As the statistics show, on Windows, 2 million out of 20 million packets were lost, or one in ten.
Uptime: 3 days.
On Proxmox:
Uptime: 19 days, 63 million packets were sent to vmbr0, 680,000 were lost.
On the physical ens13f0np0: 233 million were sent, 85,000 were lost.
It seems like packets from the Windows VM aren't reaching the physical interface and are being dropped somewhere.