Hey guys,
We have deployed logstash and kibana in our infrastructure today and everything went really smoothly with the deployment aside from the small but unexpected issue with our ES configuration. We run our ES clusters with action.auto_create_index=false (which is a recommended setting for production AFAIU) and our logstash ES server deployment uses the same setting. As far as I understand, kibana3 expects index auto-creation to be enabled and just fails on any dashboard saving operations on a 404 response from ES. We have created the index manually and it works like a charm now, though we've spent some time trying to figure out what was going on.
If you were to add an additional step to your installation docs telling the users to create that index or just make kibana handle those exceptions, that would really improve user experience for the first time kibana users that may not realize where the issue is.
Hey guys,
We have deployed logstash and kibana in our infrastructure today and everything went really smoothly with the deployment aside from the small but unexpected issue with our ES configuration. We run our ES clusters with action.auto_create_index=false (which is a recommended setting for production AFAIU) and our logstash ES server deployment uses the same setting. As far as I understand, kibana3 expects index auto-creation to be enabled and just fails on any dashboard saving operations on a 404 response from ES. We have created the index manually and it works like a charm now, though we've spent some time trying to figure out what was going on.
If you were to add an additional step to your installation docs telling the users to create that index or just make kibana handle those exceptions, that would really improve user experience for the first time kibana users that may not realize where the issue is.