0

My wired ethernet connection drops out every 10-20 minutes, and then automatically reconnects after 1-2 minutes. I've had this issue for months, and went through what feels like hundreds of posts without any solution helping me fix it. Here are some details:

  • Running Fedora, kernel 5.13.12-100.fc33
  • I've tried multiple cables. What's more, I run Windows on the same machine/different HD, and I've never had this issue on windows. So it's likely a software rather than cable/hardware issue.
  • The network icon in the bottom bar (KDE) doesn't change when the connection drops
  • I've tried various options in Advanced Network Configuration panel, like changing link negotiation settings
  • I don't control the modem/network interface ... this is a workstation at a large institution, but IT is not touching anything Linux. Especially if Windows works fine.

Here's the output of lshw:

  *-network                 
   description: Ethernet interface
   product: Ethernet Connection (5) I219-LM
   vendor: Intel Corporation
   physical id: 1f.6
   bus info: pci@0000:00:1f.6
   logical name: eno1
   version: 00
   serial: 6c:2b:59:d3:39:57
   size: 1Gbit/s
   capacity: 1Gbit/s
   width: 32 bits
   clock: 33MHz
   capabilities: pm msi bus_master cap_list ethernet physical tp 10bt 10bt-fd 100bt 100bt-fd 1000bt-fd autonegotiation
   configuration: autonegotiation=on broadcast=yes driver=e1000e driverversion=5.13.12-100.fc33.x86_64 duplex=full firmware=0.1-4 ip=134.129.67.100 latency=0 link=yes multicast=yes port=twisted pair speed=1Gbit/s
   resources: irq:93 memory:92f00000-92f1ffff

Any thoughts? Could this be a kernel issue?

[Edit to add:]

Possibly some more info -- I'm running journalctl -ef, and when it goes down, I see entries like: Sep 03 10:50:39 localhost.localdomain NetworkManager[1237]: <info> [1630684239.5911] policy: set-hostname: current hostname was changed outside NetworkManager: 'localhost.localdomain'

and

localhost.localdomain systemd-resolved[1144]: Using degraded feature set UDP instead of UDP+EDNS0 for DNS server XXX,.XXX.XX.XXX. Sep 03 11:18:45 localhost.localdomain systemd-resolved[1144]: Using degraded feature set TCP instead of UDP for DNS server XXX.XXX.XX.XXX.

7
  • I have not. TBH, I don't know which log file I'd look at in that directory. Sep 1 '21 at 16:11
  • try /var/log/syslog and /var/log/messages. is authentication needed to access network ? (wasn't able to see it on your lshw extract)
    – Archemar
    Sep 2 '21 at 5:08
  • I don't have a syslog, but grep-ing for eno1 (my ethernet interface) in messages returns a lot of entries. I'll need to sift through them more carefully, but I wonder if something like Sep 1 10:41:11 localhost NetworkManager[1285]: <info> [1630510871.4829] dhcp4 (eno1): option dhcp_lease_time => '3600' could be the culprit. Sep 2 '21 at 14:13
  • If DHCP lease time is 3600 seconds, then os should renew lease every 1800s at least. Last I used unbuntu in coroporate network, lease time was in days, maybe configure host for more frequent renewal (I can't even point you to the right direction) ?
    – Archemar
    Sep 2 '21 at 14:37
  • Possibly some more info -- I'm running journalctl -ef, and when it goes down, I see entries like: Sep 03 10:50:39 localhost.localdomain NetworkManager[1237]: <info> [1630684239.5911] policy: set-hostname: current hostname was changed outside NetworkManager: 'localhost.localdomain' Sep 3 '21 at 15:53
0

How is the ip configured (static vs dhcp)? If you know any unallocated ip, can you temporarily set the ip into that?

Could be an issue of duplicate ip within the network.

1
  • It's dhcp. I'm on an enterprise network, so I don't think I can identify unallocated IPs. I wonder if I can enable some setting that would re-allocate a new IP to me. Sep 2 '21 at 14:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.