Hi.
I am no consul developer, but I think that’s intended, because the service definition is stored locally on the node (/consul/cluster/services/) and thus an agent information. A service in the consul infrastructure catalog lives so long a node holds the service locally.
Because we talking about /v1/agent/service/register ... can you say prove my issue?:
If you are trying to do HA then you should be running more than one instance of the service on more than one node and load balancing/failover between them.
Then if node-01 goes down the load balancer won't send it requests.
This feature of consul is actually pretty nice since if your node goes down, your service will be automatically marked as unhealthy. If it worked another way, your node might go down without taking the service registration with it.
In other words, since your service is really running on that node, if the node crashes, so does the service. They are in the same failure domain.
Nicholas
--
This mailing list is governed under the HashiCorp Community Guidelines - https://www.hashicorp.com/community-guidelines.html. Behavior in violation of those guidelines may result in your removal from this mailing list.
GitHub Issues: https://github.com/hashicorp/consul/issues
IRC: #consul on Freenode
---
You received this message because you are subscribed to the Google Groups "Consul" group.
To unsubscribe from this group and stop receiving emails from it, send an email to consul-tool...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/consul-tool/89cb64b7-1f14-4a29-a310-888a15f85226%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.