When a machine (machine A) with a validator running on it is dying for some reason, I would start it on another machine (machine B). To avoid slashing, I have to copy the slashing protection history (slashing protection database) to the new machine.

For this purpose I can do regular exports of the slashing protection history on A and store them somewhere. Let's assume I do this every 5 minutes. Now, in the worst case, when I start the validators on B my history is 5 minutes outdated.

I have two questions here:

  1. Can this cause slashing to my validators?

  2. Can I also create a new history file on B that only allows the validators to start some epochs later? (missing some attestations is not a problem, but it would avoid doing regular exports on A)

  • Don't have a specific answer for you, but in general you should really not need to worry about slashing due to a period of 5 minutes. Even if there was slashing for that period, it would be extremely minuscule Aug 16 at 5:24
  • Thanks for your answer! :) I thought that if slashing happens, it means that the validator is directly forced to be deactivated. But indeed it seems that there is a separation between a punishment (not too problematic) and a slashing (validator is offline). So you would say that in my example I could only expect a very small punishment, but no slashing? (which would be acceptable ;)
    – sagacity
    Aug 19 at 8:55
  • Nobody can be forced to be deactivated. I don't remember the terms either, but even if you're considered offline (node not up to date) you can only lose a tiny amount in 5 minutes. Aug 19 at 9:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.