Hi, thank you very much for your answers, I tried but I think that a single vm with 4GB of ram is not enough to run both istances:
I copied CNODE_HOME with different configuration to be a producer and it can also works (its actually listening on 3001) as, the relay alone can work (listening on 6001), but when I try to enable both togeter the second instance is killing the first one
I think that it will be not possible to run both in a cloud instances with 4GB of RAM also because requirements says that you need at least 2 servers with at least 4GB of ram, I was tinking that I can have a single server doing both funtions and using 4GB of ram if I didnt want to separate funtions…
Then in the end I think I will need another virtual cloud instance… there is no alternative right? I need both producer and relay running as separate process and requiring me 4GB of ram…
Just to start with something it is enough to have the block producer node alone… relays needed only for security reasons. (mainly DDoS attack) That means if you start the block producer node with a topology file which was generated to a relay node (without listing the block producing node) you can start participating on block creation.
Hello,
If all you want is to experiment with testnet, run the block producer node only, no problem.
For mainnet though, set-up a relay or two. They are there to protect your bp and your money.
probably something wrong with the environment variables… one instance should not kill the other - they are two independent processes - only the system could interrupt one of them (because of out of memory or something…)
@Triton-pool I understand why putting 2 relay are fore sure better for redundancy, anti DDos and having a better service, but I can’t understand very well why having an only relay and a producer in separate machine would be so better: if one system is down all service is down despite if is a producer or a relay I think and the most possible security is to be implemented both in producer than in relay.
Thank you very much for your suggestion and for yor help.
One thing you dont need to do is to modify the topology.json - since it will be handled by topologyUpdater script. but what you dont want to put as a remote address is your node… topology file contains only remote address where your node will connect to…
Hello Andrea,
i believe there is never a 100% secure system, but one should always try to get as close as possible. In this sense spending extra $7/month for second vps is well justified with the reasons you already stated. On the top of it your bp node only need to talk to your relays, all else can be blocked including ping response. The relays are typically open to the whole internet to accomodate others nodes talking to them. To me this alone is worth the 7 bucks too.
@Triton-pool sorry asking you and I hope not to be unkind ,but where can you obtain a vps for 7$/month? The one i deployed is 20 euros/month for 4GB 80G hdd and that’s why I’m trying to optimize, maybe if not is possible in forum could you give me a suggestion via pm?
