I also have a cluster with "error" from last week. The only option that the web console is giving me is to delete the cluster. The cluster is inaccessible even from docker commands.
And today, another cluster went down.Although the web console shows the cluster as active, in the images and containers section, I only see the loading message. Now the cluster is inaccesible even from docker commands.
I know that we should use this on production environments, but so many down make this system horrible even from a dev or testing environment... can you give us a hand?