-
-
Notifications
You must be signed in to change notification settings - Fork 5.4k
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
HTTP monitors all have a status of "canceled" after update to 1.23.4 #4027
Comments
Same here. As a workaround, I noticed that enabling |
Ops, that is strange. I thought I handled it. Need to investigate. I thought #4011 was fixed issue, but clearly not. Do you guys know how to use SQLite Expert to view data? I want to see what is inside your |
|
I tried reproducing this by creating a new database in branch Maybe you can post the config of the monitor here (masking out anything sensitive), and the node.js version used (or docker). |
not for me :( |
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
Thats a weird issue! So this only happened after upgrading to 1.23.4? and should i upgrade to 1.23.4? or should i wait? |
After downgrading to 1.23.3 the monitors are up and running again. |
thank you, I'm going to do the same as I keep receiving Cancelled notifications for no reason and my paused monitors get unpaused magically :/ |
Monitors unpausing is another issue that should be fixed with #4032. Again, for those who have experienced this exact issue (HTTP monitor getting "Cancelled" when it's actually up), please post your node.js version and OS (or docker) so we can help troubleshoot. 🙇🏻♂️ |
I saw about #4032, good to know a fix is coming, thank you. My versions: |
I'm not sure if I had exactly the same issue regarding the Cancelled message. It shows repetitive Down/cancelled events: |
Just release a beta version for fixing this issue:
Would be appreciated if you guys could help to test it. |
That's strange. I just released 1.23.6. Free feel to try. If it is still not working. Please change the tag to 1.23.3. |
Installed 1.23.6 but issue is still there. So reverting to 1.23.3. Could it be related to IPv6? Pushover fails to send notifications:
Also there are services that are not affected by this strange "cancelled" bug. Both of them have only A Records. Hit me up if there is anything else I could help you with! |
@JMyrng Thank you so much. I just created a new image for printing more info and +10 seconds for AbortSignal (#4053)
Now the log will output something like this:
The monitor id could be found in the browser address bar. Would be appropriated if you could show the log here for the monitors that are having issues. |
|
Wait, it seems that it is actually a real timeout that is longer than 5 seconds? |
Thanks for your help, which make me wondering, we might be looking into a wrong direction. It may be a Node.js 18 issue. I will build two more docker images later. |
Looking forward to the results. 👀 |
This is another layered cake situation as I managed to reproduce the "Retries getting reset" part of the problem.
I think we should either backport #3072, or increase the |
You are right, the pattern is similar to @JMyrng's graph, assume that he did not restart the monitor manually. |
So i just have to restart the monitor? When i updated i completely removed the old & pulled out the new ver, i assumed that was enough, i'm using portainer tho |
No, that comment was replied to @chakflying's comment. You could try these images I mentioned in #4027 (comment) |
I'm starting to think it's my container's issue cause i went into portainer container, click on console and then pinged google.com and these were the results |
Has node 18 changed anything about the use of IPv6? My IPv4 only monitors are still not affected by this bug. I believe that IPv6 is not set up correctly in the container and the preferred variant has changed. I'll try to disable IPv6 later. |
Starting from node v17, the DNS lookup of node would preserve the IPv4/IPv6 priority returned by the resolver (nodejs/node#39987), whereas before it would always prefer IPv4. This means that compared to before, some connections will switch to using IPv6 when available. |
How can I set dns.lookup verbatim option to false to test this? |
Got it disabled in the docker network, everything is working now with 'latest' tag. So in my case it was a incorrectly configured ipv6 setup. |
@JMyrng Thanks for testing and finally found the problem. I will mention this in the changelog, as I unlikely downgrade back to Node.js 16. @chakflying For #4027 (comment), feel free to continue in a new issue. |
I have same issue, 1.23.6, sometimes got canceled status update from 1.23.3 to 1.23.6 node version 20.9.0 |
🛡️ Security Policy
Description
HTTP monitors all showing "canceled" after update to 1.23.4 and it must be related to #3961 (comment)
When I checked, the timeout for these monitors were all set to 2880 seconds (I don't think I set this). I set them down to something more reasonable (a couple of seconds) and all monitors work again.
I didn't see any other posts about this and it may just affect me as I've been running Uptime Kuma for a long time now.
FIX: Update the timeout of your HTTP monitors to a new value.
👟 Reproduction steps
Have old monitors and upgrade to 1.23.4 (from 1.23.2 in my case).
👀 Expected behavior
Monitors will report being up.
😓 Actual Behavior
Monitors reported down with the status "canceled".
🐻 Uptime-Kuma Version
1.23.4
💻 Operating System and Arch
UBuntu 22.04
🌐 Browser
N/A
🐋 Docker Version
No response
🟩 NodeJS Version
No response
📝 Relevant log output
No response
The text was updated successfully, but these errors were encountered: