#elasticsearch-kubernetes Elasticsearch (1.6.0) cluster on top of Kubernetes made easy.

This work is based on pires/kubernetes-elasticsearch-cluster. It is a merged, simpler version that serves as loadbalancer, master and data node.

Pre-requisites

• Kubernetes cluster (tested with v0.19.0 on top of Baremetal machines + CoreOS)
• kubectl configured to access your cluster master API Server

Test

Deploy

kubectl create -f service-account.yaml
kubectl create -f elasticsearch-service.yaml
kubectl create -f elasticsearch-controller.yaml

Wait until elasticsearch-controller is provisioned

Validate

I leave to you the steps to validate the provisioned pods, but first step is to wait for containers to be in RUNNING state and check the logs of the pod (as in Elasticsearch):

kubectl get pods

You should see something like this:

$ kubectl get pods
NAME                         READY     REASON    RESTARTS   AGE
elasticsearch-chlp5          1/1       Running   0          5m

Copy pod identifier and check the logs:

kubectl logs elasticsearch-chlp5

You should see something like this:

*$ kubectl logs elasticsearch-chlp5                                                                               
log4j:WARN No such property [maxBackupIndex] in org.apache.log4j.DailyRollingFileAppender.
log4j:WARN No such property [maxBackupIndex] in org.apache.log4j.DailyRollingFileAppender.
log4j:WARN No such property [maxBackupIndex] in org.apache.log4j.DailyRollingFileAppender.
[2015-07-09 17:12:50,164][WARN ][bootstrap                ] Unable to lock JVM memory (ENOMEM). This can result in part of the JVM being swapped out. Increase RLIMIT_MEMLOCK (ulimit).
[2015-07-09 17:12:50,395][INFO ][node                     ] [Shadow-Hunter] version[1.6.0], pid[1], build[cdd3ac4/2015-06-09T13:36:34Z]
[2015-07-09 17:12:50,396][INFO ][node                     ] [Shadow-Hunter] initializing ...
[2015-07-09 17:12:50,985][INFO ][plugins                  ] [Shadow-Hunter] loaded [cloud-kubernetes], sites []
[2015-07-09 17:12:51,099][INFO ][env                      ] [Shadow-Hunter] using [1] data paths, mounts [[/data (/dev/sda1)]], net usable_space [11.8gb], net total_space [15.9gb], types [btrfs]
[2015-07-09 17:12:55,362][INFO ][node                     ] [Shadow-Hunter] initialized
[2015-07-09 17:12:55,362][INFO ][node                     ] [Shadow-Hunter] starting ...
[2015-07-09 17:12:55,733][INFO ][transport                ] [Shadow-Hunter] bound_address {inet[/0:0:0:0:0:0:0:0:9300]}, publish_address {inet[/10.244.1.3:9300]}
[2015-07-09 17:12:55,771][INFO ][discovery                ] [Shadow-Hunter] elasticsearch/7cpzZ8ZPRzifvJhyls7bNg
[2015-07-09 17:13:02,464][INFO ][cluster.service          ] [Shadow-Hunter] new_master [Shadow-Hunter][7cpzZ8ZPRzifvJhyls7bNg][elasticsearch-chlp5][inet[/10.244.1.3:9300]]{master=true}, reason: zen-disco-join (elected_as_master)
[2015-07-09 17:13:02,510][INFO ][http                     ] [Shadow-Hunter] bound_address {inet[/0:0:0:0:0:0:0:0:9200]}, publish_address {inet[/10.244.1.3:9200]}
[2015-07-09 17:13:02,511][INFO ][node                     ] [Shadow-Hunter] started
[2015-07-09 17:13:02,588][INFO ][gateway                  ] [Shadow-Hunter] recovered [0] indices into cluster_state
[2015-07-09 17:14:47,080][INFO ][cluster.service          ] [Shadow-Hunter] added {[Ebenezer Laughton][ReSUZAkPQ768m1LJYKnpwg][elasticsearch-usyb5][inet[/10.244.103.5:9300]]{master=true},}, reason: zen-disco-receive(join from node[[Ebenezer Laughton][ReSUZAkPQ768m1LJYKnpwg][elasticsearch-usyb5][inet[/10.244.103.5:9300]]{master=true}])
[2015-07-09 17:14:47,259][INFO ][cluster.service          ] [Shadow-Hunter] added {[Hannah Levy][9h5NbVAyS6yrmtJiB-DocQ][elasticsearch-rqw9x][inet[/10.244.83.4:9300]]{master=true},}, reason: zen-disco-receive(join from node[[Hannah Levy][9h5NbVAyS6yrmtJiB-DocQ][elasticsearch-rqw9x][inet[/10.244.83.4:9300]]{master=true}])

Scale

Scaling is as easy as:

kubectl scale --replicas=3 rc elasticsearch

Access the service

Don't forget that services in Kubernetes are only acessible from containers in the cluster. For different behavior you should configure the creation of an external-loadbalancer, in your service. That's out of scope of this document, for now.

kubectl get service elasticsearch

You should see something like this:

$ kubectl get service
NAME            LABELS                                               SELECTOR                                             IP(S)          PORT(S)
elasticsearch   clustername=proofofconcept,component=elasticsearch   clustername=proofofconcept,component=elasticsearch   10.100.5.176   9300/TCP
                                                                                                                                         9200/TCP
kubernetes      component=apiserver,provider=kubernetes              <none>                                               10.100.0.2     443/TCP
kubernetes-ro   component=apiserver,provider=kubernetes              <none>                                               10.100.0.1     80/TCP

From any host on your cluster (that's running kube-proxy):

curl http://10.100.5.176:9200

This should be what you see:

{
  "status" : 200,
  "name" : "Termagaira",
  "cluster_name" : "elasticsearch-k8s",
  "version" : {
    "number" : "1.6.0",
    "build_hash" : "cdd3ac4dde4f69524ec0a14de3828cb95bbb86d0",
    "build_timestamp" : "2015-06-09T13:36:34Z",
    "build_snapshot" : false,
    "lucene_version" : "4.10.4"
  },
  "tagline" : "You Know, for Search"
}

Or if you want to see cluster information:

curl http://10.100.5.176:9200/_cluster/health?pretty

This should be what you see:

{
  "cluster_name" : "elasticsearch-k8s",
  "status" : "green",
  "timed_out" : false,
  "number_of_nodes" : 3,
  "number_of_data_nodes" : 1,
  "active_primary_shards" : 0,
  "active_shards" : 0,
  "relocating_shards" : 0,
  "initializing_shards" : 0,
  "unassigned_shards" : 0,
  "number_of_pending_tasks" : 0,
  "number_of_in_flight_fetch" : 0
}