ironfish
ironfish copied to clipboard
iron-fish
Feature: add rescan/decrypt progress in status output
Description
When the node become synced from unsynced state seems it continue to decrypt transactions and do "re-play" of balance changes since unsynced state to current, it do it slower than sync process. need to add indication of that process in "status" output. sure, if my assumptions is correct. the same when you doing rescan, it should show decrypt process state in percent
To clarify - I think what you're saying here is the underlying data structures that provide balances, etc lag behind the syncing process. How much so, in your experience?
Also, I would consider this a bug, or at the very least unclear UX. Changing label accordingly.
To clarify - I think what you're saying here is the underlying data structures that provide balances, etc lag behind the syncing process. How much so, in your experience?
yes, it depends on cpu. i use now zfs snapshots instead of regular rescan, before .37 it took about 3-4 hours of 3-5 days after sync is complete from rollbacked snapshot (3-5 days old) (5950x). but on last versions .40 and .41 it do decrypt even if you create new address, 2 days since upgrade, still cannot complete and i dont know state of it. and much slower than on old versions (i suppose it starts from beginning now if node not synced or some other issue causing this feeling). moreover, rescan now freezes if the node is running and do not proceed at some random point. i know that wallet 2.0 on the way, but if it far from completeness, may be do some simple things like:
- do not start rescan/decrypt for newly created address
- specify the block from which start rescan
- save in db blocks where current addresses have transactions and re-use it in next rescans
- if the address list is empty skip decrypt pos to current height.
I've been testing the wallet 2.0 branch today/yesterday and it's awesome. Rescans and balances are super fast. I don't think we need to save rescan positions of accounts, etc.. no need to complicate things.
I do think it'd be beneficial to add some clarity in the CLI response for the account status. We should probably add the scan status and the endHeader of the chainProcessor.
