Enable warp sync #901
Enable warp sync #901
Conversation
|
@shunsukew about this PR, have you tested how it works? I'm asking becausae I found this PR while doing the uplift analysis and my plan was just to create a follow-up item to implement this later, once all parts have been merged. But I guess they already are? |
yes, I'm testing it now |
|
Although Warp sync is unstable, Normal block sync works as expected. |
|
@bkchr @cheme We will conduct I see, lazy migration is ongoing just changed |
Yess that is the right infos, did a quick sumup here: paritytech/polkadot-sdk#14 Note that a "lazy migration" may never complete (if a single value of size > 32 byte is not rewritten in state which is very likely). |
I don't get this. Could you please explain what you mean by this? |
|
IIUC "lazy migration" is the hybrid mode, one can wait for all the state to be rewritten and thus in state version 1. |
|
Ahh! By "lazy migration" you meant we just wait for the entire state getting "touched". Makes sense now! Ty for the explanation. |
| // This is done for example when gap syncing and it is expected that the block after the gap | ||
| // was checked/chosen properly, e.g. by warp syncing to this block using a finality proof. | ||
| // Or when we are importing state only and can not verify the seal. | ||
| if block_import.with_state() || block_import.state_action.skip_execution_checks() { |
There was a problem hiding this comment.
https://github.com/paritytech/substrate/issues/13921#issuecomment-1518812972
This is for shiden.
| relay_chain_interface: relay_chain_interface.clone(), | ||
| }) | ||
| .await?; |
There was a problem hiding this comment.
Just interested, how come we can get rid of block_announce_validator_builder? Is it not required or deprecated?
Pull Request Summary
Enable warp sync feature for local and parachain services.
Check list