add test for a 1-block reorg in short_sync_backtrack() #18966
Conversation
arvidn
added
Changed
|
|
||
| # validate reward coins | ||
| for reward in rewards: | ||
| rec = records.pop(reward.name()) |
There was a problem hiding this comment.
what was the purpose of moving this here ?
There was a problem hiding this comment.
I remove the records from the database that has been accounted for from the actual block in order to ensure the list is empty at the end. Otherwise I couldn't detect additional coin entries in the database, that don't exist in the block.
| ) | ||
|
|
||
| print(f"chain A: {chain_a[-1].header_hash.hex()}") | ||
| await add_blocks_in_batches(chain_a[-1:], full_node_1.full_node, chain[-1].header_hash) |
There was a problem hiding this comment.
i think chain_a[-1] instead of chain_a[-1:] since the intention is to only add one block here
There was a problem hiding this comment.
but add_blocks_in_batches takes a list of blocks, not just a single one
There was a problem hiding this comment.
oh right, ok i think [chain_a[-1]] would maybe be more clear for someone who is not familiar with what we are doing
| chain_b = bt.get_consecutive_blocks( | ||
| 1, | ||
| chain, | ||
| guarantee_transaction_block=False, |
There was a problem hiding this comment.
theres a better way to do this instead of randomly trying, you can enforce the sp the block would be created after, so you can make sure the chain_a would be infused after the chain_b block, you can also farm it on a new slot entirely that would have the same effect
check out
arvidn
force-pushed
the
short-sync-backtrack-test
branch
from
e0a0f75
to
a275a82
Compare
|
arvidn
force-pushed
the
short-sync-backtrack-test
branch
from
a275a82
to
9f94652
Compare
Purpose:
Add test coverage for the fixed bug. This test fails when run on top of main, but passes with Almog's fix.
Current Behavior:
1-block reorg in
short_sync_backtrack()is not covered by a test.New Behavior:
1-block reorg in
short_sync_backtrack()is covered by a test.