In order to add redundancy, would it be advisable to run several BP/Relay combos in different data centers with all BPs having the same keys? Could it lead to problems because all BPs cloned in such a way would produce competing blocks?
It’s my understanding that the setup you’re proposing should technically work, at this moment in time, in the Cardano ecosystem. It is though discouraged by the Cardano ecosystem and pools having 2 BP active producers could be penalised in future. Unfortunately I can’t find anymore the articles where I’ve ready this.
To recap, 2 active BP could work today, but they are discouraged and possibly penalised in the future.
I hope that cardano devs will implement some sort of active/active protocol that would allow multiple redundant BP to coexist so to minimise chances of missing a block and allow redundancies for SPOs
my 2 cents.
That‘s what I thought. Having many pools systematically producing competing blocks using cloned BPs can‘t be desirable.
What is the current best practice for SPOs to ensure continuity of service during node software updates without producing duplicate blocks? I can only think of reducing the probability of missing a slot using some clever plumbing at the network layer, but that looks like an ugly hack that might not even work in all cases.
I could find no discussion about this topic nor anything in the documentation. It surely can’t be that every pool is running on a single BP, right?
There is another thread w/ some insights.
My recommendation would be to time your downtime.