-
-
Notifications
You must be signed in to change notification settings - Fork 3.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Systemd 249 network issues after waking from sleep #20244
Comments
Could you enable to generate debugging logs of networkd and provide it? The logs can be generated by creating the following drop-in config:
|
I see the same behavior with systemd The SLAAC global address is continuously being removed and added to the interface. For me it isn't 100% reproducible after every suspend to RAM cycle but it always occurs after some cycles. Another log to cross reference: |
Can also confirm this behaviour with 249 on Arch Linux. SLAAC addresses alternate between being added and removed with every incoming RA. This only starts happening some time after booting. The resume from suspend others have reported is not directly related, just often a way to move forward in time a few hours. This comment in issue #20050 references this behaviour: #20050 (comment) I would guess commit 899034b apparently did not completely fix this issue, but rather delays it until a few hours after a reboot. |
…utes configured by NDisc Fixes systemd#20244.
…utes configured by NDisc Fixes #20244.
just for reference: i've been using v249 + the proposed patch since this morning including two temporary s2ram + resumes and have not had any issues yet. looks promising! |
@foxxx0 Thank you! |
Unfortunately, after 2 days with longer s2ram periods (8-10h) the issue now presented itself even with the patch. It seems that on one RA the SLAAC works and the interface gets configured. Did a quick restart of networkd debug log: https://gist.github.com/foxxx0/ff83f6130a9c2b4c7c073a2a52076319 |
@foxxx0 From the debugging log, the first and third RA messages seem to contain autonomous prefix, but the second and fourth do not. So, the IPv6 address is once configured, but removed later. See the following:
(As you can see, there is no Could you check that the router always send information about the autonomous prefix? If so, could you also check that this behavior can be reproducible with earlier releases, e.g. plain v249 or v248? |
With systemd v248 I've had no such issues even with power-on times for over two weeks with daily suspend periods. From my ISP I have been assigned a static
output from
Please let me know if you need further information. |
Just as an update, I have now dumped 20 consecutive RAs from my router and all of them contain the same information, including the
block. So the RAs should not be the issue here but rather what systemd does with them. |
…n/update timestamp Follow-up for 25db3ae and 899034b. Fixes another issue in systemd#20244.
…n/update timestamp Follow-up for 25db3ae and 899034b. Fixes another issue in systemd#20244.
compiled, installed and rebooted. will continue to observe and report back in 2-3 days if it keeps working as intended. (or earlier if it fails before that) |
Thank you! |
…utes configured by NDisc Fixes systemd#20244. (cherry picked from commit 10e417b)
Almost reached 3d uptime with multiple s2ram periods, some of them even for ~12 hours and haven't encountered this issue yet. |
Thank you for testing! |
…n/update timestamp Follow-up for 25db3ae and 899034b. Fixes another issue in systemd#20244. (cherry picked from commit 5865dc1)
systemd version the issue has been seen with
systemd 249 (249-3-arch)
Used distribution
Arch Linux
Linux kernel version used (
uname -a
)5.12.15-arch1-1
CPU architecture issue was seen on
amd64
Expected behaviour you didn't see
Network connection to resume and stay constant with ipv6
Unexpected behaviour you saw
Frequent network connection changes with ipv6 (global) address disappearing and then reappearing. I can reproduce this problem identically with a desktop and a laptop. Rebooting fixes the problem until next sleep. Workaround is to disable systemd-networkd and systemd-resolved, and enabling NetworkManager after which network connection resumes normally even after waking from sleep.
Steps to reproduce the problem
Put the computer to sleep, and then wake it up. This problem only happens after sleep.
Additional program output to the terminal or log subsystem illustrating the issue
The text was updated successfully, but these errors were encountered: