Checklist

• My issue is specific & actionable.
• I am not suggesting a protocol enhancement.
• I have searched on the issue tracker for my issue.

Description

Right now, moving to new repo is tedious, user needs to manually:

• move pins
• move mfs
• preload local blocks to other node
• (optionally) move identity, config

Example:

• user is running legacy badgerv1 datastore, and they would like to move all their data to a fresh node
• user is running ipfs-cluster without GC, and would like manually do a spring cleaning and drop unpinned blocks by moving only the currently pinned data to the new repo.

Solution

Would be good to have CLI for this that that takes target repo path or RPC port and performs export/import between two Kubo instances.

Copy to repote RPC (online, over HTTP)

$ ipfs repo copy /ip4/127.0.0.1/tcp/5002

Copy to local repo (offline, direct filesystem access)

$ ipfs repo copy /path/to/new/repo

And ability to specify what to include/exclude (by default we don't migrate config):

$ ipfs repo copy --pins=true --mfs=true --config=false --identity=false --all-blocks=true

• pins migrates ipfs pin ls --names -t recursive (migrate pins and their name/type + preload)
• mfs migrates ipfs files ls / -l (lazy, no preload, or walk and only preload things that were already preloaded on old node)
• all-blocks migrates (preloads) all blocks from blockstore ipfs refs local
• ipns (false by default) ipns keys and last verisons of published records
• config (false by default) config TBD (all? only peering/routing/remote pinnignservices
• identity (false by default) warn that its dangerous