Ghost Ethernet Devices

MackSix

MackSix

Verified User
Joined
Apr 5, 2021
Messages
43
Techs at HiVelocity don't support DirectAdmin, but told me enp0s20f0u8u3c2, enp1s0f0, and enp1s0f1 were "ghost" ethernet devices and could cause the problem I was having with server being unresponsive and having to restart. Since the last DirectAdmin updates (including Kernel) I don't have the problem. Also, enp1s0f1 & enp0s20f0u8u3c2 is running. I don't know why they would be a "ghost" if they are running. I don't know if HiVelocity tech knew what he was talking about.

Anyhow they moved the configurations for these 3 devices from /etc/sysconfig/network-scripts/ to /etc/sysconfig/network-scripts/old and I still had the server become unresponsive until the Directadmin update.

I did find that ethernet_dev = enp1s0f0 in directadmin conf and that was showing an error in the directadmin error log file. I changed it to ethernet_dev = eno1.

Now I still show this and wonder if those 3 ghost devices are there for any reason and how do I completely remove them if not needed?

This is AlmaLinux 8.8 & DirectAdmin 1.651. Thanks!

Code: 
Last login: Wed Jul 26 14:39:45 CDT 2023 on pts/0
[admin@server ~]$ /sbin/ifconfig
eno1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 107.xxx.xxx.xxx  netmask 255.255.255.248  broadcast 107.xxx.xxx.xxx
        inet6 fe80::d250:99ff:feff:ed66  prefixlen 64  scopeid 0x20<link>
        ether d0:50:99:ff:ed:66  txqueuelen 1000  (Ethernet)
        RX packets 10534132  bytes 1968475747 (1.8 GiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 15422763  bytes 15171261014 (14.1 GiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
        device memory 0xa1300000-a137ffff

eno2: flags=4099<UP,BROADCAST,MULTICAST>  mtu 1500
        ether d0:50:99:ff:ed:67  txqueuelen 1000  (Ethernet)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 0  bytes 0 (0.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
        device memory 0xa1100000-a117ffff

enp0s20f0u8u3c2: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        ether 6a:28:8e:a8:6b:19  txqueuelen 1000  (Ethernet)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 168  bytes 27888 (27.2 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

enp1s0f0: flags=4099<UP,BROADCAST,MULTICAST>  mtu 1500
        ether a0:36:9f:51:24:ec  txqueuelen 1000  (Ethernet)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 0  bytes 0 (0.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

enp1s0f1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        ether a0:36:9f:51:24:ee  txqueuelen 1000  (Ethernet)
        RX packets 0  bytes 0 (0.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 168  bytes 27888 (27.2 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536
        inet 127.0.0.1  netmask 255.0.0.0
        inet6 ::1  prefixlen 128  scopeid 0x10<host>
        loop  txqueuelen 1000  (Local Loopback)
        RX packets 26230  bytes 8086898 (7.7 MiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 26230  bytes 8086898 (7.7 MiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
 
Last edited:
I've never seen it like this. I've also got Almalinux installations, and one has an enp2s0 as network interface, the other has eth0.
The interface name doesn't really matter.

But I never seen what you have. When looking at the "ether" statements, due to the different mac addresses it looks to me like there are multiple nics installed in the server which are just not connected, or something like that.

@Zhenyapan you seen something like this before?
 
if you have eno1 - put it to DA conf, also shutdown this three devices.
maybe you have docker or something else that can create this devices? but they didn't have IPs, strange.
 
Here is some more info. enp0s20f0u8u3c2 looks like something to do with the bios. Maybe the IPMI? enp1s0f1 looks like a real integrated nework chip that is 10GB and emo1 is 1GB. I am supposed to be on the Internet with a 10GB connection and it appears I am not.

Code: 
[root@server ~]# nmcli
eno1: connected to eno1
        "Intel I210"
        ethernet (igb), D0:50:99:FF:ED:66, hw, mtu 1500
        ip4 default
        inet4 107.155.81.82/29
        route4 107.155.81.80/29 metric 101
        route4 default via 107.155.81.81 metric 101
        inet6 fe80::d250:99ff:feff:ed66/64
        route6 fe80::/64 metric 256

enp0s20f0u8u3c2: disconnected
        "American Megatrends Virtual"
        ethernet (cdc_ether), 6A:28:8E:A8:6B:19, hw, mtu 1500

enp1s0f1: disconnected
        "Intel X540-AT2"
        1 connection available
        ethernet (ixgbe), A0:36:9F:51:24:EE, hw, mtu 1500

eno2: unavailable
        "Intel I210"
        ethernet (igb), D0:50:99:FF:ED:67, hw, mtu 1500
 
in such case just configure correct dev in DA so it can go through it to internet and check license.
 
After all of that, then the server stopped being responsive. This has been happening a lot. They acknowledged the CPU was bad and replaced the CPU. Finally!
 
It still went down with the new CPU, so they replaced the RAM... then they replaced the Motherboard. Then they hooked the WAN up to the wrong port when they replaced the motherboard. It is now finally working. ?
 
Zhenyapan said:
hope they give you at least month compensation?
Click to expand...
They never offered anything, but I expect to get the 30% discount on renewal of the server that I got before. I forgot to add, one tech disabled Kernel Care auto update and I found that Kernel Care had no license installed and wasn't updating anyway. I was paying extra for that and it wasn't even working for the last 6 months. ?
 
Last year, a HiVelocity tech gave me some total bullshit about including a firewall block file in CSF. I even had managed support then. The year before last, I paid extra for their best DDOS protection for another sever, but it didn't work and I got hit with slowloris. They had to configure rules on the server and then they overdid it and it was blocking legit traffic.
 
MackSix said:
Last year, a HiVelocity tech gave me some total bullshit about including a firewall block file in CSF. I even had managed support then. The year before last, I paid extra for their best DDOS protection for another sever, but it didn't work and I got hit with slowloris. They had to configure rules on the server and then they overdid it and it was blocking legit traffic.
Click to expand...
How much pain you are willing to take, before switching to another provider?
 
johannes said:
How much pain you are willing to take, before switching to another provider?
Click to expand...
I need a provider with "any legal content" in the USA. I don't like a TOS that says no "extremist" content or no "offending content" and other ambigous terms like that. Any recommendations via PM are welcome, if it is OK with the rules here.
 
You must log in or register to reply here.
Back
Top