01 August 2018 Releases

Elasticsearch Service: Hot-Warm Architecture, Machine Learning, and More

By Anurag Gupta

We are thrilled to announce the next evolution in the Elasticsearch Service on Elastic Cloud, which brings a ton of new features to give you the ability to solve more use cases, operate at a larger scale than ever, and give you even more capabilities to customize your deployments.

A quick look at some of the new features:

  • Manage hosted Elasticsearch and Kibana from one place
  • Add new machine learning nodes to deployments to gain insight and identify anomalies
  • Use new deployment templates backed by optimized hardware for I/O, CPU, memory, and storage-based use cases
  • Deploy hot-warm architecture with index lifecycle management
  • Scale to multiple Kibana instances with custom capacity accessing the same Elasticsearch cluster
  • Create dedicated master nodes for larger-scale deployments to improve resilience
  • New transparent pricing and a free allowance to make getting started even easier

Try it out with a free 14-day trial

All of these new features build on top of the already available features for the Elasticsearch Service for Elastic Cloud that simplify deploying, managing, and upgrading hosted Elasticsearch.

Manage Your Entire Elastic Deployment From One Place

With this new release, you can define and configure all of the Elastic Stack components supported by the Elasticsearch service from one page. Just pull the right slider and scale Kibana instances, or add some RAM to your master nodes.

Dedicated Master Nodes

We know many of our customers are solving problems on the petabyte scale and need the ability to scale their Elasticsearch clusters. That’s why we are introducing the ability to deploy dedicated master nodes that are independently scalable and configurable.

Machine Learning for All Your Deployments

Thousands of Elastic customers already use our machine learning features to simplify operations, identify anomalous behavior, and reduce time to root cause analysis. Now available with the Elasticsearch Service on Elastic Cloud, you can quickly add machine learning features to each of your deployments and start gaining insights today.

Optimized Deployments for Your Use Case

From logging to security analytics to application search, companies are using Elasticsearch to solve some of the largest search problems. With the latest release of the Elasticsearch Service, we give you the option to select underlying hardware that is optimized for your use case. We’re also introducing the concept of deployment templates to make getting started super easy.

deployment templates

Hot-Warm Deployments with Index Lifecycle Management

One of the most powerful features of the Elasticsearch Service is our hot-warm architecture template, which lets you ingest and query current data quickly while keeping older, longer-term data on denser and more cost-effective hardware. Now you can quickly deploy this architecture in minutes and set up index lifecycle management without having to modify any configuration files.

create.png

New Reduced Pricing

Along with all the new features and capabilities we are also making getting started with the Elasticsearch Service on Elastic Cloud even more accessible with a new lower starting price and a free allowance. Pricing now follows a simple linear model based on size and number of availability zones. Additionally, each deployment can now include multiple types of nodes and instances, such as Elasticsearch data nodes, Elasticsearch master nodes, and Kibana instances, with each instance priced separately.

Here are the key changes in our new pricing model:

  • A lower starting price of just $16.40 / month
  • Free tier to start gaining insights without breaking the bank
  • Straightforward linear pricing for each stack component and underlying hardware configuration  
  • New transparent snapshot storage and data transfer cost
  • Visibility into underlying cloud resources

Additionally, while in the past we bundled data transfer, snapshot storage, and cluster costs into an all-inclusive plan, we found that some customers paid more for portions they did not use. We made the pricing model much more fine-grained with independent charges for data transfer (in/out/between nodes), snapshot storage (total storage size and requests), and deployment cost (which includes cluster nodes and Kibana instances).

To help customers better understand these new data transfer and snapshot storage costs we are offering a new promotional period where these are not charged. In addition, once we start charging for snapshot storage and data transfer, we will include a set amount of storage and transfer as part of our free allowance.

New Free Allowance

One of the most exciting announcements for our deployments is a new free allowance to help you get started with the Elasticsearch Service.

The free allowance includes the following per deployment:

  • A single 1 GB Kibana instance, available for the entire life span of the deployment
  • A single 1 GB machine learning node, available for the entire life span of the deployment

Additionally, the free allowance also includes the following snapshot storage and transfer allocations per account:

  • Snapshot storage
    • 100 GB in storage across all deployment
    • 100K API requests across all deployments
  • Data transfer
    • 100 GB data transfer (In/Out/Between Nodes) across all deployments

For more information on the new pricing and the free allowance, along with frequently asked questions, please visit our pricing blog post.