Closed diego1q2w closed 4 months ago
hey @diego1q2w PR looks good. Were you able to test in localnet whether all is good and whether that really fix #4632 and #4605
Theres a few conflicts @diego1q2w so I fixed them
hey @diego1q2w PR looks good. Were you able to test in localnet whether all is good and whether that really fix #4632 and #4605
Yes i tested it. it worked.
During the previous hardfork we mad sure to update the max-rate to be at least the
min-rate
+max-rate-change
to comply with the the minimum 7% rate imposed in HIP30, for validators running under that threshold. However, with that we introduced a subtle bug under the assumption thatmax-rate-change
was a small value but in practice is not, it can go up to a 100% which will sum to a number higher than a 100% when adding the min-rate. For a handful of validators that caused the sanity checks fails when persisting the state in the hard-drive. A issue that is hard to catch as it'll only happen once per epoch, this bug explains issues like the buggy validator or the validator with undelegated funds issues. As this is a change in the way the data is stored in the state this will have to go with a hardfork, I guess we can inlcude it with the leader rotation.