Mailing List Archive

good ideas
Hi!

May be its not a fresh ideas, but i think its useful.

1. For faster sync: divide block device into some equal parts. Count
hash for each block and compare it with same block on another device. If
hash different - transfer block to another node, else count next block.

2. Compress traffic between nodes - for faster transfer. Now it's
possible with compressed virtual tunnels, but i afraid "superfluous
essence".