Federation node

#1

hi, i’m trying to do the federation nodes. when i try to pair, this error occurred. need help

ps: my server name and url different from the generated link

image
image839×545 68.5 KB

#2

Can you check the parameters you’ve provided are correct; that would be the most probable cause

#3

The parameters that you mean, is it like in the picture?

image
image649×571 24.8 KB

I had put my domain name with port and added “/api/federation” (as per video that I watched) on the end point.

However the link generated to become local.cortezaproject.org domain name instead of mine.
Is there any other prerequisite setup that I missed?
Thank you for helping.

EDIT:
The server name can be anything right? Also, on the admin email, I just put a dummy email. (I assume it not required?). The server URL is the only thing that I think needed to be correct.

#4

The name can be anything; it’s a label to identify the node.
The email was meant to point to the person responsible for the federated node for administration and communication. IIRC it can safely be omitted.

#5

Hi @syahmi ,

the link that you need to provide in the input boxes is the server address of the second node, not yours, so ie https://node2.tld/api/federation.

The Generate Federation link creates a link for your instance, that is taken from the FEDERATION_HOST variable in your .env file.

You can copy/paste the generated link and add it as a new node on the https://node2.tld instance.

You can refer to this step for more info: Corteza Federation :: Corteza Docs

#6

Hi @peter ,

Thank you for the explanation. I am now able to successfully connect both nodes for federation.
The thing that I missed was:

  1. Declaring FEDERATION_HOST in the .env file.
  2. Providing the server address of the second node in the input box. My mistake was putting the server address of the first node instead of the second node.

One more thing that becomes my concerns is, when I provide the wrong server address, my second node crashes. The thing is, it crashes to the point where my docker compose cannot be restart docker compose restart ever again. I need to docker compose down my service and up again.
If this were to happen on a production service, can you advice me what should I do to prevent my data from lost?

#7

Thanks @syahmi for the detailed response.

We’ll check on this scenario and create an issue for it. What I would advise in this case is to disable federation for the node that crashed and then restart it via docker compose.