Insisive Cloud AWS EMR

Elastic Map Reduce (EMR) from AWS provides a simple interface to manage Hadoop Cluster. Elastic Map Reduce is a great way for spinning up Hadoop clusters, run analytics jobs and tear down the cluster. While customers find the convenience handy, the cluster costs can easily run up if not managed properly. With Insisive Cloud’s mix of intelligent provisioning of OnDemand and Spot Instances for the EMR Cluster, customers can gain significant savings.

Create once and use many times

With a simplified interface, you can easily launch an EMR cluster without worrying about how many spot instances or what is the bid price. Insisive Cloud also determines whether to use spot instances or defined duration loads based on the time the jobs take.

If customers like to configure the cluster as a template, you can create as many with a single click.

How Much Can an EMR provisioning save via InsisiveCloud?

Assuming you want a cluster with 10 tasknodes and 1 master node running on r4.4xlarge in N.Virginia. By seamlessly provisioning a mix of ondemand and spot instances , you cluster will cost you
10 * r4.4xlarge spot instances = 208 ( USD pm) 1 r4.4xlarge ondemand instance = 766 (USD pm)

The total cost would be 2638 USD compared to 7660 USD if all were running on demand. That is a whopping 66% savings.

With Insisive Cloud’s intelligent spot monitoring and automated provisioning which will ensure creating new spot instances or falling back to on demand seamlessly, customers can expect capacity at continued lower costs.

But Wait: What about Spot Instances support in EMR

EMR also supports using spot instances. However, there are a few key additional advantages

  • Automated monitoring for Spot terminations, snapshot of data , private ip address , elastic ip ( if applicable). This will ensure when a spot node is terminated, the data on the node is available automatically on the new node that is provisioned. As the private ip address/persisted data is preserved, the applications will see the new machines as the same as the terminated instance.
  • Insisive Cloud’s looks at multiple options for finding the appropriate instance type compared to the terminating instance that ensures there are more spot instance pools to look at. This ensures lower price for the cluster.
  • Convenience of storing cluster information as a template that can be used for quick launch. This helps dev/ test team members be able to create clusters without having to rely on the devops teams.
  • With the rightsizing recommendations, it is easy to understand if the workers are optimally utilized

If you’re looking for an alternative to investing your own scripts and quickly save upto 60% costs, try Insisive Cloud. Getting started with InsisiveCloud is quick and simple. InsisiveCloud has been scheduling spot workloads and saving users up to 60%. Also, checkout Auto Scaling Groups, Cost Governance and Security Go on, give Insisive cloud a spin with a no-obligation full featured 14-day free trail. It takes less than 15 minutes and all the savings generated are yours to keep during the trail.