To repro:

1. Create two clusters and a tribe node that is configured to join both.
2. Attempt to use PUT to create a new index.

I'm seeing the following...in this test I first added an index manually to t2, then to t1. Then I tried to add an index to the tribe node. This is a serious problem for a Kibana dashboard configured to point to the tribe node in order to span clusters.

[2014-11-11 16:04:16,527][INFO ][tribe ] [Matsu'o Tsurayaba] [t2] adding index [testindex]
[2014-11-11 16:04:27,617][DEBUG][action.admin.indices.create] [Matsu'o Tsurayaba] observer: timeout notification from cluster service. timeout setting [30s], time since start [30s]
[2014-11-11 16:04:39,616][INFO ][tribe ] [Matsu'o Tsurayaba] [t1] adding index [testindex2]
[2014-11-11 16:04:42,288][DEBUG][action.admin.indices.create] [Matsu'o Tsurayaba] observer: timeout notification from cluster service. timeout setting [30s], time since start [30s]
[2014-11-11 16:04:48,639][DEBUG][action.admin.indices.create] [Matsu'o Tsurayaba] no known master node, scheduling a retry