Linux: Unrecoverable error {"error": "Error creating tables for master database on storagenode: migrate: database or disk is full\n\

Hello after finding out that my node was outofdate due to the change in the alpha to the beta channel and updating properly, now my node keeps restarting due to the error mentioned in the title.

It says that the disk is full but:

  1. I don’t know if it refers to the disk where the storj data is saved or to the local disk of my linux system
  2. Both disk have free space, the storj one is a 2 TB disk which I only allowed storj to use 1.7 TB allowing for overhead just in case, the system disk is much more cluttered with only 100MB free (is a spare system not needed for anything special)

What can I do to solve the problem? as said the storj installation is fresh from yesterday and the log keeps stuck in that line with the node continously restarting.


df -h should tell you which one is full

None of them are full, the storj disk has 265 GB free and the system disk has 100 MB free, that is the question is reporting full when they aren’t
I would say that the storj disk is “full” for storj as it can only use 1.7 TB out of the 2 TB and the last time the node was working 1.7 TB usage was the amount reported but nothing different, I suppose that at that point the uploads to the node are rejected and people only download or delete stuff, correct?

100MB is not enogh for executing the database migration. That is the reason for the restarts.

So the database migration is done in the system disk and not in the storj disk? I supposed that everything was taking place there as the config yaml file is there the log file etc etc.
But ok I will make some room and see what happens thank you littleskunk always helpful.

I don’t know where the database is located in your setup. It would explain the restarts. You should find the .db files in one of the docker mounts.

Another usefully command would be docker prune -a. That will clean up some space.

Ok now I understand thanks for the help. :smiley: