How to get rid of dangling indices

We are running a cluster of 3 nodes and I find in the logfiles continious repeating messages:

[2018-01-02T16:24:58,539][WARN ][o.e.g.DanglingIndicesState] [lrv171d1] [[.watches/dT22WrbgTM24J0ZuROKQ2A]] can not be imported as a dangling index, as index with same name alreadyexists in cluster metadata
[2018-01-02T16:24:58,014][WARN ][o.e.g.DanglingIndicesState] [lrv171d1] [[.triggered_watches/05F1we1RQlClDhy3jVIYXA]] can not be imported as a dangling index, as index with same name already exists in cluster metadata
[2018-01-02T16:24:58,015][WARN ][o.e.g.DanglingIndicesState] [lrv171d1] [[.monitoring-alerts-6/x7M_1l6RTTWT8ZzuI8YrRQ]] can not be imported as a dangling index, as index with same name already exists in cluster metadata


I also see a double set of triggers in Kibana>management

Extra information:

  • ES 5.6.2 with x-pack on RHEL6 Jvm 1.8.0_141
  • Before we joined the nodes to a cluster they where running as a separate node , all with there own Kibana instance. So I suspect that each kibana instance created these indices on each node.

How Can I fix this?

I already fixed the problem of the double set of triggers by

DELETE _xpack/watcher/watch/S5sJQbsTQv2j1WWsEWQQuQ_*

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.