Question regarding backup daemons. about miningcore HOT 7 CLOSED

I have been running redundant Parity daemons for a couple months without issues. Using the same wallet address. They are running on two separate machines. It is how I handle east/west coast configurations. They then peer with each other.

from miningcore.

Very legitimate question. Blocks are submitted to all daemons to get them included in the mainchain as fast and as reliable as possible. Double submissions would get rejected by the network. No problem for the pool. On the other hand payments cannot be submitted to multiple daemons as this would result in double or triple payouts - ruining the pool.

from miningcore.

Yes indeed, this is what I thought as well.

So this means that coins mined in the 2nd/3rd/etc. daemon/wallet need to be manually transferred to the first/main wallet in order for the payments to be processed?

from miningcore.

@bratan I think you can import the wallet using seeds or keys on the backup daemons so they have the same access of the same wallet?

from miningcore.

@calvintam236 - This is what I thought at first (cloning the wallet for multiple daemons), but it seems this is not supported and could/will run into problems. From what I understand the synchronization will only work up until some point.

https://bitcointalk.org/index.php?topic=1986.msg25154#msg25154

from miningcore.

@bratan Thanks for the extremly helpful link. Looks like my assumption in this regard was too optimistic.

Improvements to the daemon interface are already on top of my todo list after improvements to the REST API. Since running multiple wallet copies is not a supported scenario, one of the changes I'm planning implement is to extend the tagging mechanism already used for the Monero wallet daemon to all currencies. That way the pool would continue to function even if the wallet daemon goes down, just resulting in delayed payouts.

I'm planning to augment this with new functionality in the DaemonInterface class that will continously monitor all attached daemons, take unhealthy ones out of request processing and notify the admin about the outage.

from miningcore.

@bratan how about using same blockchain (same data folder) by multiple daemons? i think this might make more sense, and it is supported

https://monero.stackexchange.com/questions/3689/why-would-you-need-multiple-daemons-using-one-lmdb-database

from miningcore.