Connection lost when high traffic

[Luca Guerrini]

Hello!
Proxmox newbie here.
Got Proxmox 6, some windows 2016/2019 VM with virtio net driver.
When the traffic on the network is high, the connection randomly drop. I must the disable/enable the device and it restart.

I've tried everything i could find on the web, but every solution fail.

The Intel E1000 seems to work perfectly, but it's not so fast.

Can Anyone help me?
What kind of informations must I post?

 

[wolfgang]

Hi,

Got Proxmox 6, some windows 2016/2019 VM with virtio net driver.

which version do you use?

What is high traffic exactly and where(Host, Guests, Windows Guest)?

 

[Luca Guerrini]

Hi,

which version do you use?

What is high traffic exactly and where(Host, Guests, Windows Guest)?

Version: pve-manager/6.1-7/13e58d5e (running kernel: 5.3.18-1-pve)
The traffic is 50 MB/sec on a Win 2019 guest that backup on another Centos guest on another server.
Ther are no events that talk about the problem. I see in the performance monitor one spike of network activity and then it goes down. I must restart the network and all is OK.
The problem, now, is also with e1000. Today it goes down during the backup. I try with many configuration (queues, tx and rx paramenters,...) but nothing wotk.
I got 2 server HP DL380 with a 4 eth bonding LACP and 4 VLAN TAG. I use the Intel hardware, not the broadcom, that is disables
03:00.0 Ethernet controller [0200]: Broadcom Limited NetXtreme II BCM5709 Gigabit Ethernet [14e4:1639] (rev 20)
03:00.1 Ethernet controller [0200]: Broadcom Limited NetXtreme II BCM5709 Gigabit Ethernet [14e4:1639] (rev 20)
04:00.0 Ethernet controller [0200]: Broadcom Limited NetXtreme II BCM5709 Gigabit Ethernet [14e4:1639] (rev 20)
04:00.1 Ethernet controller [0200]: Broadcom Limited NetXtreme II BCM5709 Gigabit Ethernet [14e4:1639] (rev 20)
0a:00.0 Ethernet controller [0200]: Intel Corporation 82571EB Gigabit Ethernet Controller (Copper) [8086:10bc] (rev 06)
0a:00.1 Ethernet controller [0200]: Intel Corporation 82571EB Gigabit Ethernet Controller (Copper) [8086:10bc] (rev 06)
0b:00.0 Ethernet controller [0200]: Intel Corporation 82571EB Gigabit Ethernet Controller (Copper) [8086:10bc] (rev 06)
0b:00.1 Ethernet controller [0200]: Intel Corporation 82571EB Gigabit Ethernet Controller (Copper) [8086:10bc] (rev 06)


Thanks in advance for your help

 

[xtriglavx]

Hi,

please, did you resolve this issue? I have the same issue with Windows Server 2016 and E1000, which is go down every day. I am using bridge with vlan, and when it LAN adapter on Windows server goes down, there isnt big traffic. For example, there was about 200Mb/s and it was OK, but yesterday was there only 8Mb/s traffic on this ethernet and LAN adapter felt down. I needed restart LAN interface in WIndows, and then LAN adapter was UP and OK.

I have ProLiant DL360 Gen9 with 4 LAN port and I am using Proxmox:

CPU(s) 48 x Intel(R) Xeon(R) CPU E5-2673 v3 @ 2.40GHz (2 Sockets)

Kernel Version Linux 5.0.15-1-pve #1 SMP PVE 5.0.15-1 (Wed, 03 Jul 2019 10:51:57 +0200)

PVE Manager Version pve-manager/6.0-4/2a719255

 

[Luca Guerrini]

Hi,

please, did you resolve this issue? I have the same issue with Windows Server 2016 and E1000, which is go down every day. I am using bridge with vlan, and when it LAN adapter on Windows server goes down, there isnt big traffic. For example, there was about 200Mb/s and it was OK, but yesterday was there only 8Mb/s traffic on this ethernet and LAN adapter felt down. I needed restart LAN interface in WIndows, and then LAN adapter was UP and OK.

I have ProLiant DL360 Gen9 with 4 LAN port and I am using Proxmox:

CPU(s) 48 x Intel(R) Xeon(R) CPU E5-2673 v3 @ 2.40GHz (2 Sockets)

Kernel Version Linux 5.0.15-1-pve #1 SMP PVE 5.0.15-1 (Wed, 03 Jul 2019 10:51:57 +0200)

PVE Manager Version pve-manager/6.0-4/2a719255

Hi,

I think it's better if you use virtio paravirtualized driver for you network device. I solved the problem i had from months ago on SERVER 2019. I don't touch the parameters of the network but the "max RSS queue". I set the RSS queue to 1 in the network device and in the proxmox network hardware. I also set down the firewall of the network.

 

[spirit]

Hi,

please, did you resolve this issue? I have the same issue with Windows Server 2016 and E1000, which is go down every day. I am using bridge with vlan, and when it LAN adapter on Windows server goes down, there isnt big traffic. For example, there was about 200Mb/s and it was OK, but yesterday was there only 8Mb/s traffic on this ethernet and LAN adapter felt down. I needed restart LAN interface in WIndows, and then LAN adapter was UP and OK.

I have ProLiant DL360 Gen9 with 4 LAN port and I am using Proxmox:

CPU(s) 48 x Intel(R) Xeon(R) CPU E5-2673 v3 @ 2.40GHz (2 Sockets)

Kernel Version Linux 5.0.15-1-pve #1 SMP PVE 5.0.15-1 (Wed, 03 Jul 2019 10:51:57 +0200)

PVE Manager Version pve-manager/6.0-4/2a719255

maybe try to update to stable kernel 5.4, it could be a bug in vhost-net in kernel 5.0/
(and maybe qemu too, proxmox 6.0 is quite old now)

 

[tribuna]

Hello Guys,
i have the same problem.
I cant find a solution.

 

[Nuno]

I had the same problem, reinstalled again from scratch and the problem didn't go away.

Then I remembered that I had in another machine but in Windows environment some problems and reboots just because I had activated XMP in the bios to take advantage of the maximum speed of my RAM (Coursair 2x32GB CL16 3200). No overclock, just the anounced by the manufacturer

I went back to the bios, removed the XMP from the AI Overclock Tuner, switching to Auto and haven't had reboots since.

So my advice is to reset default values in the bios .

Hope this could help someone.