Missing Multicast stats on RHEL guest VMs

[Hemi]

Hello,

I have a new PVE 9.1.4 cluster set up with a number of RHEL9.7 (AlmaLinux) guests, which were previously bare-metal hosts. I noticed that once the hosts were converted to VMs, received multicast packet counts stopped being reported, i.e. the count is always 0.

Looking at tcpdump, netstat -g, and other tools indicate that multicast traffic is being received, and the host has joined all the multicast groups it should. So, this just looks to be a OS reporting problem, not a multicast packet flow problem. Below is an example stat command from a VM guest:


[root@vm ~]# ip -s link show ens18
2: ens18: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
link/ether bc:24:11:c2:f5:72 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
1522132817 6165964 0 516437 0 0

My configuration uses a bridge for the guest network. I've tried disabling "multicast_snooping" on the bridge with no success.

Environment: PVE 9.1.4

PVE host network config:

auto bond1
iface bond1 inet manual
bond-slaves ens1f0np0 ens3f0np0
bond-miimon 100
bond-mode 802.3ad
bond-xmit-hash-policy layer2+3
#guest network

auto vmbr0
iface vmbr0 inet manual
bridge-ports bond1
bridge-stp off
bridge-fd 0
#guest network bridge

Guest network device: virtio


Has anyone experience similar issues or can suggest other configuration tweaks to try to get multicast packet counts reporting again?

Thanks!

 

[bbgeek17]

Hi @Hemi ,

Since you’ve confirmed that multicast traffic is being received on both the hypervisor and inside the VM, this is most likely a statistics-reporting discrepancy between the Virtio driver and the Linux kernel.

I don’t have guidance on enabling accurate multicast counters in this setup. You may want to ask on the appropriate AlmaLinux/RHEL mailing list for insight. ServerFault or similar sites can also be good venues for generic networking/virtualization questions.

---

Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

 

[Hemi]

Thank for the reply @bbgeek17,
I too was thinking is was likely a Virtio driver issue, but your post made me check the network link stats on the PVE host (which I failed to do yet), and I noticed that the multicast packet count was 0 on the VMs "tap" interface. Now I'm thinking that might be the source of the problem rather than the client VMs network driver.

IP link stats on PVE host - tap interface has a multicast count of 0:

11: bond1: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue master vmbr0 state UP mode DEFAULT group default qlen 1000
link/ether 6c:92:cf:77:f8:30 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
136997329409 180170153 0 93053 0 7823350
TX: bytes packets errors dropped carrier collsns
940512534 7305194 0 0 0 0
12: vmbr0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000
link/ether 6c:92:cf:77:f8:30 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
4816893164 82835671 0 11074792 0 2351159
TX: bytes packets errors dropped carrier collsns
1576 16 0 0 0 0
16: tap106i0: <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 qdisc fq_codel master vmbr0 state UNKNOWN mode DEFAULT group default qlen 1000
link/ether 92:4d:8c:6b:c6:a4 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
786039915 5458013 0 0 0 0
TX: bytes packets errors dropped carrier collsns
88570562213 22380034 0 0 0 0