Node fails during sync. with writer-errors about thor

[ifavo]

@ifavo Was the node ever abnormally killed, before the error?

Not that I was aware of, it is a fresh one that was running for ~48hrs, but as always with these things, I can not be a 100% certain. I am however certain that it was running for ~24hrs without interruption before the error.

I am syncing in a screen at the moment and stopped with CTRL+C at beginning a few times, because I wanted to test a few parameters (bootnode, peers, cache, etc.).

[kgapos]

This is a known issue in LevelDB that is also present in Ethereum. Are you sure all stops were graceful? A single SIGKILL or a power outage at the wrong time will lead to this corruption.

[ifavo]

I've read that thread too, did not suggest a real solution.
I deleted the data directory and started a new sync, it is past the same point now.

Thanks!

[qianbin]

@ifavo Was the node ever abnormally killed, before the error?