Kibana does not work in 7.0.0.0rc1

Bug #1793218 reported by Eric Miller
14
This bug affects 3 people
Affects Status Importance Assigned to Milestone
kolla-ansible
Invalid
Undecided
Unassigned

Bug Description

Kolla-Ansible 7.0.0.0rc1 on CentOS 7.5 built from binary images

When enabling centralized logging, Kibana's site loads in a browser, but indicates "Unable to fetch mapping. Do you have indices matching the pattern?". This previously worked great in previous versions of Kolla Ansible. Although, my testing always used images built from source, so maybe there is an issue in the binary images?

Eric

Revision history for this message
Eric Miller (erickmiller) wrote :

I have tested this with the latest stable/rocky build of Kolla and Kolla-Ansible with no change in behavior - it appears that the configuration is broken. I don't know enough about Kibana and ElasticSearch to diagnose the problem.

Attached is a screenshot of what Kibana's interface looks like after a fresh install with enable_central_logging set to yes.

Anybody have any idea what happened between Queens and the Rocky version of Kibana?

Eric

Revision history for this message
Eric Miller (erickmiller) wrote :

Just a quick note that this is still an issue with the latest built-from-source Kolla and Kolla-Ansible today (October 1, 2018), so 7.0.0.0rc2-devXX (I think dev24 and dev17 versions, respectively).

Eric

Revision history for this message
Laurent Dumont (baconpackets) wrote :

I don't believe it's broken per say. I just had the same behavior on a Stein/Centos7 with pip Kolla 8.0.2 deployment.

You can set the index to "*" and it will load the data correctly. I was looking at the deployment and it looks like the following play should have setup the default index in Kibana.

https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/kibana/tasks/post_config.yml

I'm looking for a way to just run the Kibana role and see what the output is when it tries to do the post-config stuff.

Revision history for this message
Laurent Dumont (baconpackets) wrote :

It looks like there is a ES index at :

(kolla-venv) (openstack-admin) [root@kolla-deployer share]# curl '10.10.99.44:9200/_cat/indices?v'
health status index uuid pri rep docs.count docs.deleted store.size pri.store.size
green open .kibana LTyWl_Q_SNS7HPFAR-vDRw 1 1 3 0 52.1kb 26kb
green open flog-2019.10.12 AU-IG1w-T2ulALWrdhyn_Q 5 1 45282 0 47.5mb 23.8mb

You can use flog-* and it works properly in Kibana. I'm not sure if the date will change for the index.

Changed in kolla-ansible:
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.