Volume reports 85% usage of 500MB when real content is under 1MB?

23 days ago

Project ID: 0b73c001-d4af-45e9-93ed-6d5986a0aef5 (I can't link it for some reason in the dropdown, it's hard linked to one team only)

This project has one used volume with one SQLite file, but for some reason the dashboard reports that it is using 423 MB, gradually filling up over a month

I've SSH'ed into the service using it and it only shows the one file.

```

root@87c2cc0919b2:/app/prisma/data# ls -alh

total 528K

drwxr-xr-x 2 root root 4.0K Apr 29 17:35 .

drwxrwxr-x 1 root root 4.0K Apr 29 02:55 ..

-rw-r--r-- 1 root root 520K Apr 29 17:35 bot.db

```

This seems like a bug but is there other hidden files I'm not seeing here, and if so, how can I clear them and make sure not to accumulate them? There was a lost+found folder here that I deleted aswell.

Attachments

$20 Bounty

8 Replies

Status changed to Open Railway 23 days ago


Try redeploying your service and see if space is reclaimed for old deleted files held by open processes.


Status changed to Awaiting User Response Railway 23 days ago


darseen

Try redeploying your service and see if space is reclaimed for old deleted files held by open processes.

23 days ago

I've put out a deployment yesterday and the space was not reclaimed.


Status changed to Awaiting Railway Response Railway 23 days ago


Can you run df -h in your container to displays information about file system disk space usage, and paste it here?


darseen

Can you run `df -h` in your container to displays information about file system disk space usage, and paste it here?

23 days ago

Filesystem Size Used Avail Use% Mounted on

/dev/zd2096 434M 656K 424M 1% /app/prisma/data


snazzah

`Filesystem Size Used Avail Use% Mounted on` `/dev/zd2096 434M 656K 424M 1% /app/prisma/data`

23 days ago

The 424MB here is the allocated block size, not the actual file content, the filesystem itself is only using 656KB.


medim

The 424MB here is the allocated block size, not the actual file content, the filesystem itself is only using 656KB.

23 days ago

So the filesystem just allocated up but has the graph in the dashboard shows the block size and not what it's actually using? That's fine I guess but I would need to know if limitations will get placed due to this somehow and whether that's the intended effect of the graph. Right now it's not much to bill and the observability says $0 but I don't know if that block size would impact billing if larger sizes were involved without using all of the allocated size.


snazzah

So the filesystem just allocated up but has the graph in the dashboard shows the block size and not what it's actually using? That's fine I guess but I would need to know if limitations will get placed due to this somehow and whether that's the intended effect of the graph. Right now it's not much to bill and the observability says $0 but I don't know if that block size would impact billing if larger sizes were involved without using all of the allocated size.

Volume metrics show your disk usage in the graph, not available space. So, there is an issue with your metrics, since you verified the available space using df -h . Your service will work just fine, and won't be affected by the false metrics displayed.


darseen

Volume metrics show your disk usage in the graph, not available space. So, there is an issue with your metrics, since you verified the available space using `df -h` . Your service will work just fine, and won't be affected by the false metrics displayed.

23 days ago

Yeah I know the service works fine, and the df did confirm what's really being counted here, just want to make sure this is solved since those metrics aren't accurate and want to make sure that isn't reflected in billing or alerts in the future.


Welcome!

Sign in to your Railway account to join the conversation.

Loading...