Faster node syncing achieved!
Over the past week we have been finalizing our next release: 0.12.0. We have spent a lot of time tweaking our peer-to-peer syncing protocol and it has certainly paid off :
Syncing is now ten times faster!
This substantial improvement was achieved by modifying the previous sync strategy making it possible to sync a chain much more efficiently by focusing the sync on syncing with individual forks instead of simply syncing with individual peers.
In addition, this week we also:
- Improved the chain representation by keeping track of fork points that reduce the need for block-by-block traversals which in turn speeds up many chain operations.
- Introduced the possibility to regulate the maximum number of concurrent connections (sync > max_connections) and the number of processes accepting connections (sync > acceptors).
- Exposed more configuration previously hard-coded, (sync > connect_timeout), (sync > first_ping_timeout) and (sync > noise_hs_timeout).
- Added a payload to spend transactions.
- Introduced a database table version and a startup check to handle old versions.
- Relocated user APIs for querying block by height or hash from the internal to the external endpoint.
Here is some æternity fan art 🙂
Interested in æternity? Get in touch:
GitHub | Forum | Reddit | Telegram | Twitter | Facebook | Mail
No Comments.