24 days ago
Hi, I'm seeing a large gap between the dashboard volume usage metric and what the filesystem reports inside the container.
Dashboard: "High Volume Usage is at 78% capacity" (~380 MB of 500 MB).
The chart has read 300 MB+ since the volume was created in March and only ever goes up.
Inside the container the volume holds a single SQLite database file and nothing else:
$ df -h /data
/dev/zd5696 443652 107100 326724 25% /data
$ ls -la /data
drwx------ 2 root root 16384 Mar 22 2026 lost+found
-rw-r--r-- 1 root root 109453312 Oct 27 20:56 database.db
$ du -ah /data
16K /data/lost+found
105M /data/database.db
105M /data
So actual usage is ~107 MB of ~433 MB usable. There is no WAL file, no deleted-but-open files, and the SQLite freelist is near zero.
I tried to discard unused blocks but it is not permitted in the container:
$ fstrim -v /data
fstrim: /data: FITRIM ioctl failed: Operation not permitted
Questions:
- How is the volume usage metric measured, and why would it differ from df by ~270 MB?
- Can you trim or reconcile this volume from the host side?
- The docs say a volume at 100% is taken offline for an integrity check and the service restarts. Does that trigger off the dashboard
metric or the actual filesystem usage? I want to know whether I'm at real risk of a restart when the filesystem is a quarter full.
Thanks.
1 Reply
Status changed to Awaiting Railway Response Railway • 24 days ago
24 days ago
The dashboard figure is the storage allocated for the volume on our side, not the live data your filesystem reports. Space freed inside the container only returns once it is discarded, and storage is allocated in 16 KB blocks while ext4 frees in 4 KB blocks, so partially freed blocks stay allocated. The metric settles somewhere above df rather than matching it, and your data is fine. Volume billing uses this same metric, so the roughly 0.27 GB gap works out to about $0.04/month at $0.15/GB-month.
The fstrim error is expected since containers lack the permission to trim, and we don't run manual host-side trims or reclaims on volumes.
On the 100% behavior, we can't confirm from here which of the two figures that check keys on. Right now neither is near full: the dashboard reads 78% and your filesystem reads 25%.
Status changed to Awaiting User Response Railway • 24 days ago
17 days ago
This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!
Status changed to Solved Railway • 17 days ago