-
Notifications
You must be signed in to change notification settings - Fork 1.8k
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
v0.107.31 stops responding to DNS requests, does not terminate #5896
Comments
Could you enable the verbose logging and see if there are any clues there? Enabling debug profiling ( |
Unfortunately I don't have enough space on my router for verbose logging and I'm also worried about flash wear. |
Have you checked that it's not a memory issue? If you're on a ER12, that only has 1GB of RAM. It's conceivable that once RAM is used up, it goes to Would you be able to show the As another aside, is the router still behaving as expected? Or is the CPU bottleneck causing router issues too? |
It is not a memory problem, the memory use is normally < 300MB (right now 185MB). The router is fine when this happens, except 1 CPU core is at 100% taken by AGH. Under normal use the CPU cores are at < 5%. The AGH web interface works, AGH just stops responding to DNS requests. I don't need to reboot the router to recover, I just need to kill AGH.
uptime: I've reverted to .29 for now because that's been stable for me. Problems started with the recent .30 |
I am having a simiular random issue and am running the latest v0.107.32 on centos-release-7-9.2009.1.el7.centos.x86_64. AGH stops responding to DNS requests, and still able to access the GUI. Restarting the AdGuardHome service does not work i.e. have to force a reboot. Was not able to check CPU/Memory usage yet, but never had issues before the last 1-2 releases. I did observe the logs below, and have increaed the UDP receiver buffer as described here (https://github.com/quic-go/quic-go/wiki/UDP-Receive-Buffer-Size). Am waiting to see if the issue stops... [root@dns1 ~]# systemctl status AdGuardHome Jun 16 17:18:11 dns1 AdGuardHome[734]: 2023/06/16 17:18:11.470190 [error] dnsproxy: unpacking udp packet: dns: buffer size too small |
Did not solve the issue...CPU is at 100%, enough free memory. Reverted back to v0.107.29. |
@copacetic82, those errors are likely signalling that some plain-UDP clients are sending invalid DNS packets. We'd appreciate if you could reproduce the situation with verbose logging and profile collecting, as described in #5896 (comment), if you decide to give it another go. |
Hi...thanks for replying...unfortunately i cant reproduce it at the moment as I am away on vacation. But a revert to v0.107.29 did solve the issue for me...no problems since, and using the same clients. Maybe I have time to debug once I get back. |
It's happening to me since the last update. |
Same issue. |
I face the same problem. Never had issues with AdguardHome until a few weeks. I agree this issue must have started with the last releases. |
We are still not able to reproduce this properly, and the data we've received so far unfortunately doesn't shed any light on what is happening and why. Anyone still having issues, please help us collect the data by doing the following:
|
Hello same problem this morning . whith my adguard docker version: v0.108.0-a.586+2069eddf Web UI is working This problem is during since 1 month !!! |
#5962 Had the same problem here on 0.107.33 have reverted back to 0.107.29 as suggested. |
Have set this up and left it running. |
Thanks to the logs sent by @bigwookie in #5962, we were able to pinpoint a bug in the code which could cause such issues. We've fixed it in version |
@ainar-g I have downloaded this and left it running. |
@ainar-g I can confirm it's running better. |
I've been having the same issue I believe and managed to sort out how to update to the latest Edge release over the weekend. It stalled again last night (running v0.108.0-a.600+c4651647) and I was able to grab and send a log of the event. Will follow up if it happens again since I neglected to enable the pprof option this time. |
@ainar-g Just failed again will get the logs and send on. |
just to confirm, i am also getting the same on v0.107.33
A reboot of both fixes is temporarily LXC ADH set to Primary DNS and Docker is Secondary in DHCP Ive had to switch the LXC Container off as ive no idea how to downgrade it or even use the Edge channel, but ive set the Docker to use Edge and its pulled v0.108.0-a.600+c4651647. il monitor it |
Yes and yes. |
Version |
I am now running Version: |
For those who follow the Beta channel, v0.108.0-b.40 contains the same fixes. |
Nice Work @ainar-g |
@bigwookie How does the new version work |
So far it is running smoothly with no crashes or errors as before. @ainar-g thank you for this, it's looking good. |
The fixes have now been released to v0.107.34. Once again, thanks to @bigwookie, @TheCableGuy99, and everyone else who tried Edge and Beta builds! |
That looks to be HomeAssistant? |
I'm having some error messages frequently in the docker image log using v0.107.34: |
Unrelated! Open your own thread! |
hello the problem is it resolved. |
v0.107.34 looks good 👍 |
Yes v0.107.34 looks good from my end, it has been running efficiently and without error since the release. |
WE Can use no edge version now ? |
Could toi explain basically what was the problem ? |
Prerequisites
I have checked the Wiki and Discussions and found no answer
I have searched other issues and found no duplicates
I want to report a bug and not ask a question
Operating system type
Other (please mention in the description)
CPU architecture
64-bit MIPS
Installation
Other (please mention in the description)
Setup
On a router, DHCP is handled by the router
AdGuard Home version
v0.107.31
Description
v0.107.31 running on EdgeRouter 12, mips64_softfloat
Sometimes it just hangs, it doesn't respond to DNS requests anymore, 1 CPU is at 100% and it also does not gracefully restart via
systemctl restart
, it needs to be force killed.I know this is not a lot to go on, these random hangs are hard to troubleshoot. It was stable for a while, I believe this started with the updates to .30 and .31
The text was updated successfully, but these errors were encountered: