tegansnyder/Upgrading Elastic Search to 2.3 - Steps I took.md

## Upgrading Elastic Search to 2.3 - Steps I took.md

      
    Raw
  

              Upgrading Elastic Search to 2.3 - Steps I took.md
            
          
    Details of steps I took on each node in a nine node ES cluster on RHEL7. These steps need to occur one node at a time, i.e. "rolling upgrade")
Step 1: Disable shard allocation

curl -XPUT "http://localhost:9200/_cluster/settings" -d'
{
  "transient": {
    "cluster.routing.allocation.enable": "none"
  }
}'
Step 2: Stop non-essential indexing and perform a synced flush (Optional)

curl -XPOST "http://localhost:9200/_flush/synced"
Step 3: Stop and upgrade a single node

# stop elasticsearch and kibana
sudo service elasticsearch stop
sudo service kibana stop

cd /usr/share/elasticsearch/

# remove plugins
sudo bin/plugin remove license
sudo bin/plugin remove kopf
sudo bin/plugin remove marvel-agent

# download RPM
cd ~/
wget https://download.elasticsearch.org/elasticsearch/release/org/elasticsearch/distribution/rpm/elasticsearch/2.3.0/elasticsearch-2.3.0.rpm

# upgrade using RPM
rpm -U elasticsearch-2.3.0.rpm

# install plguins
cd /usr/share/elasticsearch/
sudo bin/plugin install license
sudo bin/plugin install marvel-agent

# reload and restart elasticsearch and kibana
sudo systemctl daemon-reload
sudo service elasticsearch start
sudo service kibana start
Step 4: Start the upgraded node

Confirm that started node joins the cluster
curl -XGET "http://localhost:9200/_cat/nodes"
Step 5: Re-enable shard allocation

Once the node has joined the cluster, reenable shard allocation to start using the node
curl -XPUT "http://localhost:9200/_cluster/settings" -d'
{
  "transient": {
    "cluster.routing.allocation.enable": "all"
  }
}'
Step 6: Wait for the node to recover

curl -XGET "http://localhost:9200/_cat/health"

Updating Kibana to 4.5

sudo -s
service kibana stop
cp /opt/kibana/config/kibana.yml /home/myuser
cd /home/myuser
wget https://download.elastic.co/kibana/kibana/kibana-4.5.0-linux-x64.tar.gz
tar xzvf kibana-4.5.0-linux-x64.tar.gz
cd kibana-4.5.0-linux-x64
yes | cp -R * /opt/kibana/
yes | cp -R /home/myuser/kibana.yml /opt/kibana/config/kibana.yml
cd /opt/kibana/
chown -R kibana:kibana *
service kibana start
Installing plugins using Ansible:

I added the Centrify dzdo privilege escalation method to Ansible in PR ansible/ansible#15219
# install graph plugin
ansible elk -m shell -a '/usr/share/elasticsearch/bin/plugin install graph' --become --become-method=dzdo
ansible elk -m shell -a '/opt/kibana/bin/kibana plugin --install elasticsearch/graph/latest' --become --become-method=dzdo

# restart elasticsearch
ansible elk -m service -a "name=elasticsearch state=restarted" --become --become-method=dzdo

# restart kibana
ansible elk -m service -a "name=kibana state=restarted" --become --become-method=dzdo
You can use sudo if you are not using Centrify dzdo. Your /etc/ansible/hosts/whatever file would look like this:
[elk]
esmaster01.mydomain.com
esmaster02.mydomain.com
esmaster03.mydomain.com
esclient01.mydomain.com
esclient02.mydomain.com
esclient03.mydomain.com
esdata01.mydomain.com
esdata02.mydomain.com
esdata03.mydomain.com

[elk]
esmaster01.mydomain.com
esmaster02.mydomain.com
esmaster03.mydomain.com
esclient01.mydomain.com
esclient02.mydomain.com
esclient03.mydomain.com
esdata01.mydomain.com
esdata02.mydomain.com
esdata03.mydomain.com

[elk:vars]
ansible_ssh_user=mysshusernamehere
public_key_file=/Users/myusername/.ssh/id_rsa.pub
become_method=dzdo