Managing several instances of a canister on a network

Hello all,

What is the best practice for managing multiple deployments of one canister on a single network?

Historically what I have done is to create several “networks” in the dfx.json networks section but that actually all point to the same network (local dfx server or mainnet) but I guess this isn’t a good idea long term.

At one stage I had several different checkouts of the same repo, so each had its own canister_ids.json, which worked but was very cumbersome.

How do you do it?

The best thing we have is the multiple networks that point to the same thing (we call it staging network in the docs). With 0.12.0 of dfx with the system-wide replica this will be a lot less cumbersome because you can describe the multiple networks in one place and access them from everywhere, but the base problem will still be the same.

1 Like