-
Notifications
You must be signed in to change notification settings - Fork 529
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Don't check ledger hash at start; restore final checkpoint #13600
Conversation
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
!ci-build-me |
The fix for payment test hasn't been merged for compatible. There's some merge conflict hasn't been resolved with the berkeley and develop branch (the test executive module has been changed). Helena is picking up the work this week. My suggestion for now would be just retry the test manually. |
!ci-build-me |
After loading the blocks it's going to replay, the replayer builds a table from global slots to ledger and state hashes. If the start slot did not contain a block, it would search backwards and never find the most recent hashes, because only the blocks to be replayed are loaded, no blocks before the start slot. So don't look for the ledger hash at the start slot -- we haven't yet replayed anything. We didn't see this issue until we started using checkpoint files.
Also, restore the checkpoint file dumped at the finish, but only write a checkpoint file if the block at a slot is canonical.
Updated the cron job to use image from this PR.