Everything deleted on my disk

Node Operators troubleshooting
,
1

Hello, I have a problem. Yesterday I deployed my 1st node, I used docker on RPi4. I used empty 4GB disk, I copied my identity there. So in the docker compose I have something like:

  volumes:
      - type: bind
        source: /mnt/str/identity/storjnode1
        target: /app/identity
      - type: bind
        source: /mnt/str/
        target: /app/config

After running 1st docker with SETUP=true I saw several directories created on the disk.
I started the node and it reported OK on dashboard, 4 satelites 100% all, network going up.
In the morning I was invited with emails from storj about problems. My docker container reported error about missing identity. I checked - disk was empty. Just directory where previously was identity and empty directory bin.

What I did wrong? What happened? I can of course create node again, but repeating my errors would not lead me anywhere… Thanks!

2

Hello @hwm.land,
Welcome to the forum!

I suppose you mean 4TB disk.

This is mean that either you disk is dismounted, or it doesn’t have a partitions table and you formatted the disk without creating a partition.
I would suggest to check that the disk is still mounted:

df --si -T

If it’s mounted, but empty, I would suggest to repartition it, as described in the guide:

Or use an LVM to create a logical volume, then format it with ext4.
You may see an example there:

It’s also possible that you just need to check it and fix errors with fsck.

3

Thank you for your help. Yes, it’s 4TB, just typo. And you are right, disk is not mounted anymore. Didn’t occuret to me that this could happen and I’ll still see directories with ls.
Any idea why this could happen? It’s USB3 connected drive, but there were chia files for ~2 years with no issue at all. Strange enough - I was not able to sudo restart the system (it did not start, I have to power cycle) After this node started again. After power cycle I’ll not find anything, but how to investigate it should it happen ever again? (I’m just beginner in linux)

4

Have you got the mount point added to your /etc/fstab?

Also, you say you’ve been using chia for 2 years. Was this on another OS? If so, why make the change to Linux now (just curious)?

5

Yes, sure.

It was on the same machine (RPi4), the same OS, as well in container. Disk was always ext4 statically mounted. I just deleted all (chia) files.

6

USB drives are not designed for 24/7 operation, with a large number of input/output operations the USB controller can overheat and turn off the drive, also the drive must have its own power source, RPi cannot provide the necessary power under load.

7

The difference between Chia and Storj disk activity is night and day, by an exponential factor. Your USB setup is probably beyond it’s stress tolerance. Consider new equipment.

1/2 cent,
Julio

8

Hello Alexey, Julio, thank you for all your effort to help me!

Yes, I know. I have good 12V+5V power supply for my 2 HD drives and one SSD. System serves me as home-server for years running ~5 small containers for my needs (personal wiki, git… nothing big nor havily used, but handy)

There is an installation article on the official storj site about installing on RPi3 (so only USB2) with no world of warning, so I did not expected issues on RPi4, seems I was wrong. Well, I’m anyway in the process of building new home server on X86, but for few months RPi must hold strong :neutral_face:

9

They are designed a little bit different, and I mean the controller in the USB case, not on your RPi. You are correct, there is no warning because you basically have no choice :smiley:
I used a power supply, the case and the extension board from GeekWorm, they are pretty good and the disk is worked fine.

10

Also, adding to Alexey’s points, don’t forget not all USB cables are created equally.

Good luck,
Julio

1 Like