Compressed filesystem missing in node_filesystem_size_bytes and node_filesystem_free_bytes

[werdck]

Host operating system: output of uname -a

Linux andavir 5.10.0-11-amd64 #1 SMP Debian 5.10.92-1 (2022-01-18) x86_64 GNU/Linux

node_exporter version: output of node_exporter --version

node_exporter, version 1.1.2+ds (branch: debian/sid, revision: 1.1.2+ds-2.1)
  build user:       [email protected]
  build date:       20210725-21:22:06
  go version:       go1.15.9
  platform:         linux/amd64

node_exporter command line flags

None

Are you running node_exporter in Docker?

No

What did you do that produced an error?

Fetched node_filesystem_size_bytes

What did you expect to see?

All the partitions on my host and their sizes

What did you see instead?

The expected output minus one partition

Info about the partition:

$ mount | grep sdb1
/dev/sdb1 on /mnt/archiv type btrfs (rw,relatime,compress=zstd:15,space_cache,subvolid=5,subvol=/)

df -h /dev/sdb1
Filesystem      Size  Used Avail Use% Mounted on
/dev/sdb1        37T   20T   17T  54% /mnt/archiv

# btrfs subvolume show /mnt/archiv
/
	Name: 			<FS_TREE>
	UUID: 			<UUID>
	Parent UUID: 		-
	Received UUID: 		-
	Creation time: 		2021-11-22 16:20:49 +0100
	Subvolume ID: 		5
	Generation: 		39095
	Gen at creation: 	0
	Parent ID: 		0
	Top level ID: 		0
	Flags: 			-
	Snapshot(s):

[werdck]

Note: might be the issue #1848 reappearing

[ventifus]

I see this most commonly when the node_exporter user doesn't have permission to access the mount point (/mnt/archiv in this scenario). Try running df as the node_exporter user and see if you can see the mount point. On my systems, it runs as "prometheus", so I would do sudo -u prometheus df, please adapt to the specifics of your system.

[werdck]

Thanks for the hint. The user (also prometheus in my case) can see the partition in df just fine.