We are running docker swarm in production. 4 node cluster.
Docker-compose starts the applications.
We are using consul as the discovery (k/v) backend (clustered consul, 4 nodes)
With this setup we are seeing excessive calls to the /docker/network/xxx - keys in consul.
We see up to 40 requests per second to consul, many of them targeting the same resource.
We suspect that all this traffic makes our consul cluster unstable.
Are there any known limitations or bugs making this happen?
Are there any options for caching network information