Cluster start timeout error for ElasticSearch 6.2.2

[wahlg]

Running Elasticsearch::Extensions::Test::Cluster.start always gives me a timeout error Timeout while waiting for cluster status [green] and [2] nodes. I've tried different numbers of nodes and increasing the timeout.

It seems like the same issue as #371, but I'm using a later version of ElasticSearch (6.2.2 instead of 5.x).

[wahlg]

It looks like this was happening because WebMock requests to localhost were disabled. So all the requests to http://localhost:9250/_cluster/health?wait_for_status=green were failing. These issues weren't visible until I enabled DEBUG mode. Setting WebMock to allow localhost for all fixed the issue:

  config.before(:all) do
    WebMock.disable_net_connect! :allow_localhost => true
  end

[twp88]

Hate to be that guy, but I'm also getting the same error, and in my case it's not because of WebMock;
Timeout while waiting for cluster status [green] and [2] nodes
Timeout::Error (execution expired)

I've tried this in a variety of different ways, including changing the number of nodes, but all to no avail.

As above, seems very similar to 371, but in my case I'm running elasticsearch 6.2.3 with ruby 2.5.0 and rails 5.0.6.

[karmi]

Hmm hmm, wondering... With only a single node, the cluster should be green, I'm afraid this might be something like a bootstrap check failing or something like that. Can you run the cluster with the debug option (eg. DEBUG=yes as environment variable), and have a look at the output? It should tell what's wrong with the cluster...

[jarongao]

We're running into this issue with an update to Ruby 2.4.4, and this fixed our problem: chrisk/fakeweb#57