Would really appreciate the ability to configure multiple back ends. We have multiple data centres and each has its own Logstash and ES clusters. It would be a huge advantage to be able to have single Kibana instance be able to query across all our ES clusters and then correlate the data for display on its dashboards.
+1. We will have an ES cluster + Logstash in AWS east-1 and west-2. Would be nice to have a single Kibana accessing both.
+1
+1
+1
+1
+1
+1
Is this on the roadmap? it seems a lot of people including myself would love to see this feature.
Assembling the data across disparate clusters is really not possible. Facets couldn't count correctly, statistics could not be calculated, sorting wouldn't be possible. I'm leaving the issue open because maybe some day Elasticsearch will have a mechanism for dealing with it, but its not something Kibana can do.
Any idea of how hard it would be to modify kibana for a limited implementation? I'm looking at creating several sites in different regions to be used for logging and doing basic search queries across them. Sorting would have to be done in ES and after the fact, but that should be a solvable problem, I'd think. In my case, I don't care about counts or statistics, when I do, it's more likely to be for a report instead of online. The other option I have thought of is a proxy that would send the query to multiple servers and aggregate the results but I don't know enough about Kibana to know which would be simpler.
I would suggest then having an independent "proxy" that can aggregate
queries across multiple ES instances...
Curtis Ruck
Anytime: 210-857-1126
On Wed, Sep 25, 2013 at 2:40 PM, Ralph Trickey [email protected]:
Any idea of how hard it would be to modify kibana for a limited
implementation? I'm looking at creating several sites in different regions
to be used for logging and doing basic search queries across them. Sorting
would have to be done in ES and after the fact, but that should be a
solvable problem, I'd think. In my case, I don't care about counts or
statistics, when I do, it's more likely to be for a report instead of
online. The other option I have thought of is a proxy that would send the
query to multiple servers and aggregate the results but I don't know enough
about Kibana to know which would be simpler.—
Reply to this email directly or view it on GitHubhttps://github.com/elasticsearch/kibana/issues/22#issuecomment-25113332
.
that's more on the elasticsearch side, cause we have all the data. Tricky, but not imoossible....
OK, thanks. I'll head down that road then. If I can get a successful start, I'll try to get permission to set it up as open source.
+1
tribe node feature was added in elasticsearch that would enable it. In Kibana, we just miss multi index wildcard support which should be in soonish
+1
Is there a ES version which would let us play with both Kibana 3 and Kibana 4?
We would like to build some dashboards with the same underlying data to allow users to feedback on which Kibana version they prefer.Thank you.
@mish2015 I am using Kibana 3.1.0 and Kibana 4.1.1 on top of elasticsearch 1.4.5 and it works fine with out any issues.
I guess it possible because both old and new kibana are using different indexes to store its objects
@ZbigniewZabost-zanox -Thank you, i will see if we can do the same, would make the comparison way easier.
+1
This functionality should be available by using a "Tribe" node, see:
https://www.elastic.co/guide/en/elasticsearch/reference/current/modules-tribe.html
Is there any way we can use single Kibana to access multiple AWS elastic search clusters? I understand tribe node is one way but AWS elastic search doesn't work with tribe nodes.
Most helpful comment
+1. We will have an ES cluster + Logstash in AWS east-1 and west-2. Would be nice to have a single Kibana accessing both.