Resync of validator gets stuck with wrong Block.Header.AppHash error #2860
Description
Expected Behavior
When we have a valid validator that proposed blocks in the past, dashmate stop --platform && dashmate reset --platform && dashmate start --platform should resync platform on the node up to the tip.
Current Behavior
We got multiple reports that resync sometimes fails with apphash mismatch error.
ERROR: failed to start node: error on replay: wrong Block.Header.AppHash at state height XXXXX, block XXXXX. Expected AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA, got BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB
Looks like this happens on a height where that node was a proposer.
Possible Solution
Steps to Reproduce (for bugs)
On a validator that proposed blocks long time in the past:
dashmate stop --platform && dashmate reset --platform && dashmate start --platform