OpenSearch in a Nutshell: A Fun and Informal Guide

The History: How We Got Here

Once upon a time, there was a search engine called Elasticsearch. It was fast, scalable, and loved by many. But then, the evil corporate overlords (aka “That Company That Shall Not Be Named But Starts with an A”) decided to take Elasticsearch in a more… business-y direction.

Enter OpenSearch—the people’s champion! It was born when Amazon forked Elasticsearch 7.10.2 and said, “We shall make this open-source again!” And thus, in 2021, OpenSearch came to life, free from licensing restrictions and corporate greed.

Since then, OpenSearch has been on a mission to provide an open-source, community-driven search and analytics suite. It’s like Elasticsearch, but with more freedom and fewer angry emails about licensing violations.

Why OpenSearch?

Why should you care about OpenSearch? Good question! Here’s why:

• It’s Free and Open-Source – No sneaky licensing shenanigans!
• Compatible with Elasticsearch – You can use most Elasticsearch tools and APIs with OpenSearch.
• Great for Logs, Analytics, and Searching – It’s not just a search engine; it’s a full-fledged analytics powerhouse.
• Scalable and Fast – Can handle tons of data across distributed nodes.
• Kibana? Nope! Meet OpenSearch Dashboards – A friendly UI for visualizing all your data needs.

Now that we’ve covered the why, let’s get our hands dirty with some code!

Getting Started with OpenSearch

Installing OpenSearch

You can get OpenSearch running locally using Docker:

1
2
3
4

docker run -d --name opensearch -p 9200:9200 -p 9600:9600 \
  -e "discovery.type=single-node" \
  -e "OPENSEARCH_INITIAL_ADMIN_PASSWORD=admin" \
  opensearchproject/opensearch:latest

This will spin up an OpenSearch node on http://localhost:9200.

Now let’s make sure it’s working:

1

curl -u admin:admin -X GET "http://localhost:9200"

If everything goes well, you’ll see some JSON with details about your OpenSearch cluster.

Indexing Data in OpenSearch

Let’s add some data! Say we have a collection of cats (because why not?).

1
2
3
4
5

curl -X POST "http://localhost:9200/cats/_doc/1" -H "Content-Type: application/json" -u admin:admin -d '{
  "name": "Whiskers",
  "age": 3,
  "color": "gray"
}'

Boom! You just added a cat to OpenSearch. Congratulations!

Searching for Data

Now, let’s search for our furry friend:

1

curl -X GET "http://localhost:9200/cats/_search?q=name:Whiskers" -u admin:admin

And there you have it! Your OpenSearch cluster is now home to a fine feline.

More Advanced OpenSearch Stuff

Using OpenSearch Dashboards

OpenSearch Dashboards is the graphical interface for OpenSearch. You can run it using Docker as well:

1
2
3

docker run -d --name opensearch-dashboards -p 5601:5601 \
  -e "OPENSEARCH_HOSTS=[\"http://localhost:9200\"]" \
  opensearchproject/opensearch-dashboards:latest

Then, open http://localhost:5601 and start exploring your data visually!

Aggregations (Because Data Needs Stats)

Want to count how many cats you have by color? Easy!

1
2
3
4
5
6
7
8

curl -X POST "http://localhost:9200/cats/_search" -H "Content-Type: application/json" -u admin:admin -d '{
  "size": 0,
  "aggs": {
    "cats_by_color": {
      "terms": { "field": "color.keyword" }
    }
  }
}'

This gives you a breakdown of how many cats you have per color. Maybe it’s time to diversify your cat collection?

Wrapping It Up

So, OpenSearch is basically Elasticsearch’s cool, open-source sibling. It’s powerful, scalable, and free. If you’re dealing with logs, analytics, or search-heavy applications, OpenSearch is a fantastic choice.

And hey, if you ever get lost, just remember: curl is your best friend.

Key Ideas

References

• OpenSearch Official Docs
• OpenSearch GitHub
• Docker OpenSearch