Annotation query failed on Internet dashboard in Grafana

[duindain]

Thx for making this, I've just installed today on Rasp Pi OS 32bit on a raspberry pi 3 clean install

When opening Grafana Internet connection dashboard i always see an error

Annotation Query Failed
e is undefined

I haven't customised Grafana from the default install, do you know how to resolve this please?

The numbers look correct for my connection but the charts seem to not be working other than the number display

Cheers

[duindain]

Just an update all the graphs are rendering correctly now after some time has passed, however every time the dashboard is loaded/refreshed it still says the original error e is undefined

[Filadeus]

Hi! I also did a clean install of Rasbian OS (x32) on to Rasberry Pi 4 (4Gb) today, and encountered a same issue. Alert box says Annotation Query Failed Cannot read property 'to' of undefined every time the dashboard updates.

[johntdavis84]

Seeing this as well on an existing install.

It appears to be related to updates to grafana or node-exporter.

I have Watchtower installed on my Docker host, and got a message to update both of these packages. As soon as I did, I started seeing the error.

Here's the hashes for the new images:

Found new grafana/grafana:latest image (4947919fd4d6)
Found new prom/node-exporter:latest image (f0789a50d794)

[unapproachable]

Here to say 'me too'. Deployed another version last week and it doesn't display the error/warning, but on the version I deployed two days ago, I also get the error about annotations. Any recommendations/pointers greatly appreciated before I go an learn the whole of Grafana, Ansible, and Docker

[geerlingguy]

I'm guessing the latest version of Grafana needs an update to the dashboard. I'll mark this as a bug and investigate later if I can, then export the updated version of the dashboard. I haven't updated my Grafana container image in a few weeks.

[unapproachable]

For those being driven mad by this, as you excitedly refresh this page to see how your ISP is doing, you can do the following to correct it while waiting for Grafana to fix their biz:

Be sure to edit the deployed version of the files in the "internet-monitor" directory, not the ones in the internet-pi/internet-monitor project folder. Should be at the same level as the internet-pi and pi-hole (if deployed) directories.

Specficially, edit internet-monitor/docker-compose.yml changing line 93 from:

image: grafana/grafana

to

image: grafana/grafana:8.0.6

Then, rebuild the container:

docker-compose up -d --no-deps

Output should be something like this:

Pulling grafana (grafana/grafana:8.0.6)...
8.0.6: Pulling from grafana/grafana
e160e00eb35d: Already exists
0dedfb7c9040: Pull complete
88be4fbe8ef0: Pull complete
4f4fb700ef54: Pull complete
2441661bc981: Pull complete
11bed60a1ce4: Pull complete
a349d3a4a214: Pull complete
Digest: sha256:66c120542437c0d31a76e8370322e6bfcc5c8c78041d9acf9eaa3e2b40f0fe6b
Status: Downloaded newer image for grafana/grafana:8.0.6
internet-monitoring_ping_1 is up-to-date
internet-monitoring_speedtest_1 is up-to-date
internet-monitoring_nodeexp_1 is up-to-date
internet-monitoring_prometheus_1 is up-to-date
Recreating internet-monitoring_grafana_1 ... done

This specifies the version of Grafana to use and seems to be the first one I could find going back (latest is 8.1.0) that wasn't affected. Without the version specification, we're all just pulling the lastest release, which is why the experience differs between deployments as time goes on.

Please advise if there's a better/more correct way to address this for the tool chain used here. I'm new to all of it. And, as with all things free on the Internet, this advice is worth what you paid for it.

[johntdavis84]

A new version of Grafana dropped today. I installed it and am still getting the same error. It’s no big deal—I’m almost positive it doesn’t impact functionality at all. It does distract me from how poor Reddit’s ping response is, though. :P - JTD.

…

On Aug 9, 2021, at 5:45 PM, Jeff Geerling ***@***.***> wrote: I'm guessing the latest version of Grafana needs an update to the dashboard. I'll mark this as a bug and investigate later if I can, then export the updated version of the dashboard. I haven't updated my Grafana container image in a few weeks. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#192 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGI5CYWFJ2INX7B622OASE3T4BLCFANCNFSM5BYKTLEQ>. Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email>.

[roberthunt]

Looks like the issue is caused by this bug in Grafana: grafana/grafana#37819 which breaks the "statusmap" panel used to show the uptime of the various URLs - no fix version as of yet.

There is also an issue on the statusmap panel plugin repository: flant/grafana-statusmap#219

[phdoerfler]

@unapproachable The line image:grafana/grafana:8.0.6 causes:

pi@raspberrypi:~/internet-monitoring $ docker-compose up -d --no-deps
ERROR: yaml.scanner.ScannerError: mapping values are not allowed here
  in "./docker-compose.yml", line 35, column 12

It's missing a space after the first :.

image: grafana/grafana:8.0.6 works fine.

[unapproachable]

@phdoerfler, Thanks for the catch! I've updated my original comment such that anyone copy/pasting won't have to read any more than necessary. I haven't updated anything since that change, but hopefully someone is monitoring what's going on over at Grafana with respect to this issue.

[rhyspy]

I had the same issue. I applied @phdoerfler 's fix and the problem is solved. Thank you.