# Elasticsearch

Elasticsearch is a search and analytics engine, commonly used as a substitute for Solr or for collecting log and metrics data.

You can easily add it to your Lando app by adding an entry to the services top-level config in your Landofile.

# Supported versions

• 7
• 7.9.x
• 7.8.x
• 7.7.x
• 7.6.x
• 7.5.x
• 7.4.x
• 7.3.x
• 6 (default)
• 6.8.x
• 6.7.x
• 6.6.x
• 6.5.x
• 5
• custom

# Patch versions

To use a patch version, you can do something as shown below:

services:
  myservice:
    type: elasticsearch:5.6.15

But make sure you use one of the available patch tags for the underlying image we are using.

# Other requirements

Elasticsearch requires you set the kernel property vm.max_map_count to at least 262144 in order to run correctly. Lando will attempt to set this on install if you use any of our official package installers for Windows, macOS, Debian, RPM or Pacman.

However, if you are installing from source or via another pathway you will need to set this manually. Generally this can be accomplished with the following:

sysctl -w vm.max_map_count=262144

But we recommend you check out the following issue for more information: https://github.com/lando/lando/issues/1967

# Configuration

Here are the configuration options, set to the default values, for this service. If you are unsure about where this goes or what this means we highly recommend scanning the services documentation to get a good handle on how the magicks work.

Also note that options, in addition to the build steps and overrides that are available to every service, are shown below:

services:
  myservice:
    type: elasticsearch:6
    portforward: false
    mem: 1025m
    plugins: []
    config:
      server: SEE BELOW

# Port forwarding

portforward will allow you to access this service externally by assigning a port directly on your host's localhost. Note that portforward can be set to either true or a specific port but we highly recommend you set it to true unless you have pretty good knowledge of how port assignment works or you have a very compelling reason for needing a locked down port.

portforward: true will prevent inevitable port collisions and provide greater reliability and stability across Lando apps. That said, one downside of portforward: true is that Docker will assign a different port every time you restart your application. You can read more about accessing services externally over here.

tl;dr

Recommended

services:
  myservice:
    type: elasticsearch
    portforward: true

Not recommended

services:
  myservice:
    type: elasticsearch
    portforward: 9200

# Using a custom elasticsearch.yml

You may need to override the default config with your own elasticsearch config file. Note that according to the underlying upstream image this will completely replace the default config. Further note that by default our elasticsearch services start as data nodes. If you want to activate your node to also be an ingest node then check out this example.

If you do this, you must use a file that exists inside your application and express it relative to your project root as shown below:

A hypothetical project

Note that you can put your configuration files anywhere inside your application directory. We use a config directory in the below example but you can call it whatever you want such as .lando.

./
|-- config
   |-- elasticsearch.yml
|-- .lando.yml

Landofile's elastic config

services:
  myservice:
    type: elasticsearch
    config:
      server: config/elasticsearch.yml

# Getting information

You can get connection and credential information about your elasticsearch instance by running lando info. It may also be worth checking out our accessing services externally guide.