Note keeps restarting because of dbs

I tried to fix the NTFS with my windows PC several times. It always says all is fine. But I will try to backup and reformat now.

It takes a long time to copy 10 terabytes of data with rsync. In my case about two weeks…it’s still copying. Something still seems strange to me. I actually never had problems with ntfs on linux and StorJ. But in the last weeks/months there have been more and more problems. Meanwhile I have 3 independent nodes which are stuck in the reboot loop now. Apparently it is indeed the file system. StorJ seems to do something that corrupts it recurrently.

The more space is used - then worse NTFS works without a defragmentation, but you can perform it only under Windows.
To reduce amount of copying loops you may reduce the allocated space below the usage, it will serve only egress in this case, and you will need much less iterations of rsync.

1 Like

Another node just died. If I connect the HDD to windows it says the filesystem is broken. I let it repair and defrag. All fine then, can read/write big data, no problem. Then when I plug it back to the raspi still all fine. Reade/Write working, no problem. But when I start StorJ…after a few seconds…restart loop…aaaaand ntfs broken again…how does it manage to break it again in just a few seconds?

Because the implementation of NTFS under Linux is not fully compatible, especially if the disk has been used under Windows 10 or newer. All new Windows uses extended features of NTFS, such as dedupe and compressing by default (you need to explicitly disable them, but this would be applied only to a newer content, the old one would not be touched), the Linux version doesn’t support them out of the box without compilation from the source code. So, thus I recommend to do not use NTFS under Linux at all, especially if the disk was touched by Windows 10 or newer.
You need to backup data from this disk, reformat it to ext4 and restore data. Unfortunately there is no other way.

Copy is almost done now for at least one node. I have found out something else on the other nodes. Apparently windows does not defrag everything and some things not at all. Windows also tells me that it cannot defrag the MFT (master file table) for whatever reason. I just used an external software and tried to defrag it. I hope it will save them for at least a few weeks so I can finish the copy for every node.