Is piece cache synchronization downloading data from a node or from other farms?

When adding a new SSD, sometimes the piece cache synchronization can be completed in just a few minutes when the network is good, but other times it becomes very slow, taking several hours. I want to understand the principle behind this. Where does it download these caches from? Is there a way to make it download from a location closer to me?

Piece cache sync

It primarily downloads pieces from other peers on the network, but depending on situation can download first segment from node and some pieces from plot cache.

There is no direct way to make it download from location closer to you, I feel like you’ve asked this exact question many, many times already in different ways.

If your farm is big enough, e.g. >8TB, there’ll be no piece cache sync when you add more disk/plot.

If your farm is small, it will download more piece cache to the new disk/plot.

Either way (your farm is big or small) this is same rule applied to everyone and it always help your plot time better (if farm is small) or no impact (if farm is big). And while plotting is in queue during piece cache sync, you’re actually farming with all the plotted sectors.

Maybe you’ve known all my sharing above. Just a note in case other member want a bit more clarification.

I understand you’re from China, the chinese subspace community is big enough. So even though you can’t control the peer of your node/farmer, I believe a majority is from China also. It is just the fact that piece sync is slow for me, and you, and anyone else if we don’t have good peer connections.

I am wondering that there are so many peers on the Internet. How does subspace-farmer find the quickest/nearest peer via DSN and then do the piece cache sync?

I strongly recommend you to read Introduction | Autonomys Academy for such questions, there is a lot of good information there