You are not logged in.
I have been using NetworkManager for a while, until I decided to switch to connman, which was able to keep my connection up for days without problems (I am also using a VPN service via OpenVPN manually). However, I am not able to configure connman to connect to eduroam. So, I decided to switch back to NetworkManager, which does not have problems upon connecting to eduroam. However, the connection keeps dropping every 10 to 30 minutes for no apparent reason. openvpn output does not seem to detect anything. Here are some information:
Output of sudo lshw -C network for wireless card:
*-network
description: Wireless interface
product: Wireless 8265 / 8275
vendor: Intel Corporation
physical id: 0
bus info: pci@0000:01:00.0
logical name: wlp1s0
version: 78
serial: 00:21:6b:dd:9e:3d
width: 64 bits
clock: 33MHz
capabilities: pm msi pciexpress bus_master cap_list ethernet physical wireless
configuration: broadcast=yes driver=iwlwifi driverversion=5.5.5-arch1-1 firmware=36.952d9faa.0 ip=192.168.1.3 latency=0 link=yes multicast=yes wireless=IEEE 802.11
resources: irq:138 memory:b6200000-b6201fff
Output of systemctl list-unit-files --state=enabled:
UNIT FILE STATE
autovt@.service enabled
bluetooth.service enabled
dbus-org.bluez.service enabled
dbus-org.freedesktop.nm-dispatcher.service enabled
display-manager.service enabled
docker.service enabled
getty@.service enabled
lightdm.service enabled
lm_sensors.service enabled
mariadb.service enabled
mongodb.service enabled
mysqld.service enabled
NetworkManager-dispatcher.service enabled
NetworkManager-wait-online.service enabled
NetworkManager.service enabled
postgresql.service enabled
remote-fs.target enabled
17 unit files listed.
As you can see, there is no colliding network service running apart from NetworkManager itself.
Here is a complete log by sudo journalctl -b, reporting four connection drops.
Last edited by alogim (2020-03-03 16:14:40)
Offline
Offline
systemctl list-unit-files --state=enabled
I forgot to mention that I stopped, disabled and uninstalled every existing network service/program. In fact, the output of the above command is:
UNIT FILE STATE
autovt@.service enabled
bluetooth.service enabled
dbus-org.bluez.service enabled
dbus-org.freedesktop.nm-dispatcher.service enabled
display-manager.service enabled
docker.service enabled
getty@.service enabled
lightdm.service enabled
lm_sensors.service enabled
mariadb.service enabled
mongodb.service enabled
mysqld.service enabled
NetworkManager-dispatcher.service enabled
NetworkManager-wait-online.service enabled
NetworkManager.service enabled
postgresql.service enabled
remote-fs.target enabled
17 unit files listed.
Last edited by alogim (2020-03-03 14:00:58)
Offline
Please post a complete system journal ("sudo journalctl -b") covering some drops.
Online
Please post a complete system journal ("sudo journalctl -b") covering some drops.
I edited my original question with the log from the system journal covering three drops.
Offline
No.
The complete journal of the entire boot, not fractions or filters. Eg. the 2nd journal (the only one I bothered to look at) starts w/ a dhcp error but one cannot see the context/previous events.
You can feed the journal directly into a pastebin service: https://wiki.archlinux.org/index.php/Pastebin
Guesstimates:
https://wiki.archlinux.org/index.php/Ne … HCP_client
https://wiki.archlinux.org/index.php/Ne … domization
Online
Fair enough, I edited my original question with a link to the log on pastebin.com. Hope it might prove useful to determine the cause of this problem.
Offline
Does it happen w/o OpenVPN?
Edit: to reason on this, you get a lease on wlps0, tun0 activates, 30 minutes later you get kicked (lease expires?) and the re-lease fails.
If OpenVPN is not the offender, I'd try my former two suggestions to see whether the AP objects to the MAC randomization or this is incompliant behavior by the internal dhcp client (which codebase recently changed drastically and has seen several bug reports, albeit rather reg. complete failure)
Last edited by seth (2020-03-03 21:40:04)
Online
Does it happen w/o OpenVPN?
Edit: to reason on this, you get a lease on wlps0, tun0 activates, 30 minutes later you get kicked (lease expires?) and the re-lease fails.
If OpenVPN is not the offender, I'd try my former two suggestions to see whether the AP objects to the MAC randomization or this is incompliant behavior by the internal dhcp client (which codebase recently changed drastically and has seen several bug reports, albeit rather reg. complete failure)
Well, let's explain a little. I have a script that I run which first disables all incoming and outcoming traffic with iptables and then I pick a server and enable traffic to/from that server. When I use connman and connect with OpenVPN, I can stay connected without a single moment of connection loss for the entire day. With NetworkManager, instead... well, openvpn log does not say anything (that is, it does not appear to be a problem with it, since the output stays the same even when NetworkManager does its black magic, that is, when I lose the connection). OpenVPN does not provide any useful output message in this sense, not immediately at least: after quite some time it tells me that it was not able to connect, probably when it tries to renew the lease, but with connman the lease reneweal is painless, that is, I do not "feel" it.
For example, I am know trying raw wpa_supplicant: it connects perfectly fine, then I start dhcpcd on the wlp1s0 and finally start openvpn. I am able to surf the Internet as long as I do not stop my VPN connection (of course). To be honest, it looks like NetworkManager disconnects and reconnects: however, since openvpn is started manually, this connection drop is enough for it, so I have to stop openvpn and restart it. But I would really like to know why this happens only on NetworkManager.
Offline