Elasticsearch Clustering

Registered by Swann Croiset

Elasticsearch is used to index and store logs and notifications of OpenStack deployment(s). Currently, the plugin elasticsearch_kibana deploys the solution on a single node. In order to be production ready the plugin must be able to deploy an Elasticsearch cluster offering a high availability and a scalable platform

Blueprint information

Patrick Petit
Swann Croiset
Swann Croiset
Series goal:
Accepted for 0.9
Milestone target:
milestone icon 0.9.0
Started by
Swann Croiset
Completed by
Simon Pasquier

Related branches



Gerrit topic: https://review.openstack.org/#q,topic:bp/elasticsearch-clustering,n,z

Addressed by: https://review.openstack.org/257426
    Make idempotent the importation of dashboards

Addressed by: https://review.openstack.org/257427
    Deploys Elasticsearch cluster part 1

Addressed by: https://review.openstack.org/257428
    Allow to deploy 5 Elasticsearch nodes

Addressed by: https://review.openstack.org/257429
    Add advanced settings for Elasticsearch deployment

Addressed by: https://review.openstack.org/257431
    Update index mapping for logs

Addressed by: https://review.openstack.org/257432
    Remove useless Puppet module

Addressed by: https://review.openstack.org/257433
    Upgrade Elasticsearch version to 1.7.3

Addressed by: https://review.openstack.org/257968
    Wait cluster green status before kibana dashboard import

Addressed by: https://review.openstack.org/257969
    Handle number_of_replicate parameter

Addressed by: https://review.openstack.org/258580
    Configure Haproxy for Elasticsearch and Kibana

Addressed by: https://review.openstack.org/260062
    Use the VIP of the Elasticsearch cluster

Addressed by: https://review.openstack.org/261300
    Reconfigure the cluster at scaling time

Addressed by: https://review.openstack.org/262431
    Recommend 3 Elasticsearch nodes

Addressed by: https://review.openstack.org/262452
    Set appropriately the Pacemaker option no-quorum-policy

Addressed by: https://review.openstack.org/262455
    Add into a new Monitoring group the UI settings

Addressed by: https://review.openstack.org/262793
    Specify the number of expected nodes

Addressed by: https://review.openstack.org/262794
    Configure correctly recover_after_nodes with 1 node

Addressed by: https://review.openstack.org/263324
    Use the hiera default mechanism for plugins

Addressed by: https://review.openstack.org/264169
    Use stable/8.0 for fuel-library

Addressed by: https://review.openstack.org/264170
    [WIP] Improve cluster deployment

Addressed by: https://review.openstack.org/264171
    [WIP] Initialize Elasticsearch data only once

Gerrit topic: https://review.openstack.org/#q,topic:elasticsearch-simple-monitoring,n,z

Addressed by: https://review.openstack.org/265167
    Configure Elasticsearch cluster health alarms

Addressed by: https://review.openstack.org/264823
    Monitor Elasticsearch cluster

Addressed by: https://review.openstack.org/265168
    Configure Elasticsearch node filesystem alarms

Addressed by: https://review.openstack.org/265817
    Re-execute cluster configuration on redeployment

Addressed by: https://review.openstack.org/269555
    Use the Elasticsearch VIP to configure Curator

Addressed by: https://review.openstack.org/269606
    Spread out by one hour the Curator run


Work Items

This blueprint contains Public information 
Everyone can see this information.


No subscribers.