Redundant connecction

Question

I'm tryng to build a network of computer and multichain based system.

i start first server:

multichaind xyzchain -daemon

answer from server:

Other nodes can connect to this node using:        multichaind xyzchain @1.1.1.1:4765

i go to server number 2, and i connect this server to origin server.

 multichaind xyzchain @1.1.1.1:4765

i become answer:

Other nodes can connect to this node using:        multichaind xyzchain@2.2.2.2:4336

i go to server number 3, and i connect this server to origin server.

 multichaind xyzchain @1.1.1.1:4765

i become answer:

Other nodes can connect to this node using:        multichaind xyzchain@3.3.3.3:4336

and i can set more and more servers on same system... null problems

at this we have origin server and some "first ring" servers...

then i connect "second ring of servers":

i go to server number 20 , and i connect this server to server number 2.

 multichaind xyzchain @2.2.2.2:4336

i become answer:

Other nodes can connect to this node using:        multichaind xyzchain@20.20.20.20:4337

i go to server number 21 , and i connect this server to server number 3.

 multichaind xyzchain @3.3.3.3:4336

i become answer:

Other nodes can connect to this node using:        multichaind xyzchain@21.21.21.21:4337

and so on ... null problem

if one of the "first server ring" goes down the connected server also goes down... (example: server number 2  goes down, the server 21 goes also down) .
I try to connect the server 21 to server 2 and the server 20 to server 3. in this manner i can prevent lost connection and/or brocken servers....but if i try to do this vonnection i become: answer i can't do thath connection... node alreaddy connected.

how i can connect: server A to server B & server B to server C & Server C to Server A like a triangle ?????? when i try to made last connection C to A the system answer connection alreaddy exist.

sorry for duplicated explanations.... the question is how i can interconnect a multichain servers for prevent to have isolated service servers.

Regards.

Answer 1

For a smallish network like this, the nodes in the network will all automatically discover and connect to each other, so long as there is not a firewall preventing it. So you do not need to manually create the additional connections, they will happen automatically. You can use the getpeerinfo API command to confirm this.

Comments

all server are in separate network one in London, one in New York, one in Minsk, one in Rome and so on... when one server goes down all sub-server remain isolated... how can discover the p2p connections over all internet? and if the Rome server goes down how can all sub-servers connected and granted for this server self-made a connection on the Origin server?

I have to connect all the schools.
the city's schools connect to a collector server called "city collector server". In a province, all "city collector servers" connect to a "province collector server". In a region connect "all province servers" and all "province collector servers" connect to the origin server.
I hope to have clarified that if a "collector server region", for whatever reason (the computer explodes, the disk breaks, a regional router is struck by lightning, a central internet connection is sucked by a tornado or I do not know which other reason can not connect to its superior server) that zone and all the underlying servers will not be online

sorry for my english if you don't understand tell me and i made a picture example.

---

Each node reports to the other nodes about the nodes it has discovered – that is how the word quickly gets around to all of the nodes about all of the others. In a small network, every node will therefore connect to every other one, unless firewalls prevent it.

---

First of all thanks for your answer.
I probably did not explain well ...
we speak of 35-40 thousand computers spread across a nation.
One of the 32,273 schools (Eg located on Dallas), has its own node with its assets etc etc .. but some other assets and data reads them / writes on the top node and some on the origin server (Eg: Name, Surname birth and place of birth are data that will never vary for a person and this data wil be writed onto origin server.).
Since we are talking about a network that I do not call small since 40 thousand computers are not "the baker's office" but a large network, and bearing in mind that one server is in Dallas, one in New York, one in London and one in Tokio .... I strongly doubt that if I do not tell the x server to connect both to the Tokyo node and to the node of London the nodes are self-evident in the global internet network .... maybe they will do it "in the baker's office" certainly not in the network infrastructure common as it is the Internet .... If I want to connect the computer X in London let me do but if I try to connect it to Tokyo tells me that the connection is already present it follows that all the linked sub-nodes to the X server falling the London connection, they will remain isolated and that does not suit me.
If I have to pay to get the answer is not a problem but to continue to say that in a small network does not serve does not make sense.
If multichain can't have redundant connection, i must to find another chain, while i must have al server and services 99,99 % UP and granted and perfectly working.
Regards

---

OK, I understand. So if it's a small network, the nodes will indeed all connect to each other (unless there are firewall problems). In a large network, it is not possible for every node to connect to every other, but each node will connect to a lot of other nodes, based on the maxconnections runtime parameter. So you will still end up with a fully connected network. If two peers are not immediately connected it does not matter, because transactions and blocks will be relayed between their peers.