-
Notifications
You must be signed in to change notification settings - Fork 2.4k
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
node_filesystem_device_error in v0.15.0 yet v0.14.0 works #711
Comments
One more thing to note, with v0.15.0 this symptom is happening with all GCE PD mounts in our kube clusters (even on our Google Container-Optimized OS instances).
|
Thanks for the detailed report! Would be most interested in the log output of the node-exporter. Can you provide that too? |
Sure, here are the logs for v0.15.0
In contrast, v0.14.0 has fewer permission denied errors:
|
Thanks! This looks like the same as #703, lets discuss over there. |
Okay so the root cause is the same: We're dropping privileges to nobody now, which means you can't statsfs() the bind-mounted mountpoints. If you want to monitor the mongodb-backups volume you should mount it to the node-exporter pod. Then you also need to ignore all of these failing mountpoints, e.g:
If you don't want to mount the volume to the node-exporter pod, you can of course also force it to run as root like described here: #599 (comment) |
Host operating system: output of
uname -a
Linux mongodb-backups 3.16.0-4-amd64 #1 SMP Debian 3.16.43-2+deb8u5 (2017-09-19) x86_64 GNU/Linux
node_exporter version: output of
node_exporter --version
node_exporter command line flags
Are you running node_exporter in Docker?
yes, with the recommended args
What did you do that produced an error?
Trying to access exported disk metrics for k8s attached/mounted GCE persistent disk shows a node-exporter error metric incremented when node-exporter experienced an error while collecting statistics for this device.
What did you expect to see?
When using v0.14.0, I can see the filesystem metrics for that device/mountpoint (seen below). Yet the new version v0.15.0 does not export the same metrics.
What did you see instead?
A metric implying that node-exporter encountered an error while collecting statistics for this device.
The text was updated successfully, but these errors were encountered: