-
Joshua Gutow authored
This includes an op-e2e action test which can be used as a regression test against the invalid epoch advancement. I also had to fix several other tests. This bug is caused by the batch queue advancing to the next epoch prior to having the full sequence window for that epoch. This occurs when auto generating batches. Once it has generated all of the batches for the current epoch, it creates the first batch of the next epoch before it should. Note: I had to remove the assumption that the epoch of the safe head is the same as the epoch inside the batch queue. This is not true when generating the last batch of an epoch. We can advance the batch queue, but we don't generate a safehead with the new origin.
77137e38
Name |
Last commit
|
Last update |
---|---|---|
.. | ||
derive | ||
driver | ||
sync | ||
output_root.go | ||
types.go | ||
types_test.go |