I’ve been experimenting with daml and canton. I’ve went through the tutorial DAML SDK and Canton.
please correct me if I’m wrong, as a ledger I was using DAML ledger how do I use the HLF as a ledger?
In HLF we have peers for chaincode related operations. I believe in canton participant node is helping us achieve the same thing. so how do I host different participant nodes on different servers?
Hi there!
In Canton, several participant nodes connect to a single domain. The domain coordinates the participants, for example communication between participant nodes always goes via the domain.
The domain can integrate with HLF or other ledgers. The participant nodes do not integrate with other ledgers, instead each participant node provides a consistent view and the same set of operations across all ledger integrations.
The version of Canton available now uses a PostgreSQL domain. There is a version of the domain that uses HLF under development, but it is not public right now.
You can run multiple participant nodes on different servers, and they will be able to talk to each other through the domain. For example, you don’t need to run different participant nodes in the same process or on the same host.
I hope this answers your question. Thank you for using the forum, please add any additional questions you’ve got 
Hi @Phoebe_Nichols,
Thanks for replying!
a lot of things got clear now.
I’m a little confused about the starting node part. do I use the same conf file which has information about all entities or should I break it down to the participant level? i.e separate conf file for every participant and concerned domain and use that file for starting the node.
Have a look at the example configurations in examples/03-advanced-configuration.
There you have one example with separate files for participant1, participant2 and a domain. Normally, all these nodes run in separate processes (as they are operated by separate organisations), however for convenience, prototyping and testing, we allow you to configure as many nodes in a process as you want. So yes, the configuration you have been using (simple example) would have to be broken up into separate files.
If you look at participant1-init.canton, you’ll see the commands that you need to run in order to connect a participant to the domain (so you need to run these commands on both participants).
You can run these commands interactively or running it was --bootstrap script.
I hope this helps.