Linkcast

Cast by webthe3rd.eth

webthe3rd.eth

I just validated a Snapchain storage optimization on mainnet! My patched node completed a full snapshot restore on a 1 TB Hetzner volume, peaking at just 775.7 GiB. That's ~50% less peak disk usage than an unpatched restore, potentially making independent nodes significantly cheaper to operate. PR:

fix: reclaim snapshot staging per shard instead of after all shards by webthethird · Pull Request #1046 · farcasterxyz/snapchain

Problem remove_dir_all(snapshot_dir) sits outside the loop over shards, so every shard's compressed chunks and its assembled tar stay on disk until the whole restore finishes. Peak usage is a m...

2 likes0 recasts0 replies

0 replies

No replies yet

895 link casts in the last 24 hours

Farcaster data from Hypersnap · About