thanks for the swift reply. Does it mean that I will need Contabo providing 3 VPS servers: 1x VPS plan L, 2x Relay1 VPS plan M. I was wondering if 2 servers may suffice, having 2 Relays on one VPS with 2 IP addresses: no way ? Additionally having 2 Relay nodes instead of 1 in your experience can provide higher rewards ? Thanks
Not recommended to run 2 relays on same node… in case of a crash both relays will be down and the BP will be isolated… and running more relays will help for block propagation time and the rewards will be the same (rewards are per block)
And the most difficult part is to attract delegators because u must know that the pool need ~1M ada delegated/stake for ~1block/epoch average… without this amount u will create blocks very very very seldom
Cheers,
thanks for your frankness. Cheers
In the topology and topologyUpdater files are we supposed to put the nodes’ public or private IPs? I know the producer’s not supposed to be connected to the internet, but just its LAN so I’m making sure.
if the servers are communicating via private LAN, then yes, use the private IPs
Can all nodes run on the same port (different VMs) or is it more recommendable for all of them to be different? And does the pool domain have to be accessible from the internet or just the relays’ public IPs?
Can all nodes run on the same port (different VMs) or is it more recommendable for all of them to be different?
It doesn’t matter, can run on same port or different port
And does the pool domain have to be accessible from the internet or just the relays’ public IPs?
Only the relay will “talk”/connect with other public nodes
So I don’t need a domain name unless I’m marketing my pool?
exactly, you don’t need a domain name
Pardon the ‘stupid questions’ but how would nodes outside my LAN find mine if my public IPs aren’t in the topology files? Or does CNTools handle that automatically?
the relay will use private IP to connect with BP and public IP to connect with the public network
Are the testnet nodes supposed to take a while to sync? Like several hours remaining. Speedtest gives me 243 Mbps down and 15 Mbps up.
well it depends… can take more hours
Actually I think I’m already synced because I noticed that when I started another node after a day of nonuse it synced fast then it got to Syncing: 62% and got much slower but probably because it’s waiting for the current epoch to finish.
but… are u on testnet right? what is the current sync epoch?
One node says 175 and the other 154. Yea testnet. Both running on port 6000 on two NAT KVMs on a Debian host.
then you must wait more time, I see that pools from testnet are running on epoch 205 now
Current epoch is more likely at something like 99% and nodes do not need to wait for an epoch to finish. They just sync up to the current slot and that is 100% sync finished.
On mainnet, syncing later parts has become much, much, much slower, because there is more traffic on the chain. You can easily spend half a day waiting for a few percent in the 90s. Maybe, something similar is also on testnet?
Progress is extremely slow though. Last change I made was modifying one of the node’s hostnames with hostnamectl then updating /etc/hosts then restarting cnode but I guess in a few hours I’ll update you. HeptaSean, probably. Still strange that my nodes are on different epochs though and progressing at the same slow pace.
If you started the sync at different times or paused at different times or … it’s normal. It is the epoch currently syncing.