{"id":8056,"date":"2019-02-25T10:37:34","date_gmt":"2019-02-25T10:37:34","guid":{"rendered":"https:\/\/rax-stg-wp-aws.com\/blogs\/dramatically-cut-costs-with-aws-instance-scheduler\/"},"modified":"2026-11-03T19:29:15","modified_gmt":"2026-11-03T19:29:15","slug":"dramatically-cut-costs-with-aws-instance-scheduler","status":"publish","type":"blogs","link":"https:\/\/www.rackspace.com\/en-gb\/blogs\/dramatically-cut-costs-with-aws-instance-scheduler\/","title":{"rendered":"Dramatically Cut Costs with AWS Instance Scheduler"},"content":{"rendered":"

Introduction<\/strong><\/p>\n

On-demand infrastructure, with its speed, agility, efficient use of resources and lower costs drives many organizations toward cloud adoption.<\/p>\n

When used in conjunction with tools like CloudFormation or Terraform, users are able to provision and remove cloud infrastructure  programmatically. This is typically referred to as Infrastructure as Code, or IaC, and is great for stateless resources.<\/p>\n

However, what if some servers cannot be regularly re-provisioned from scratch? Does it mean they need to be up and running 24\/7, even if only used during limited hours?<\/p>\n

Fortunately not.<\/p>\n

Cloud providers like AWS offer the ability to stop and start instances as needed while preserving the data on these instances. A stopped instance doesn\u2019t consume compute, network or power resources, so monthly costs are slashed drastically as compute charges cease (of course, storage\/disk will still be charged). This is useful for environments that are only needed for certain portions of the day.<\/p>\n

Turning off a non-production environment at the end of the day, for example, reduces weekly utilization from 168 to 50 hours, yielding a 70 percent reduction in running costs. Users can apply the same technique to proof-of-concept environments, typically used for occasional demos, or nightly batch\/job processing environments.<\/p>\n

Bottom line: not every resource or environment is needed all the time, so why pay for them 24\/7\/365?<\/p>\n

The challenge<\/strong><\/p>\n

There is a catch: automation\u2026or more precisely, the lack thereof.<\/p>\n

With the exception of auto-scaling groups, AWS doesn\u2019t have the built-in functionality to stop and start instances automatically<\/em> based on a defined schedule. You can stop your instances manually every evening, and start them back up every morning, but this quickly becomes a pain point.<\/p>\n

While some organizations build scheduling automation through services like Lambda, the maintenance can be problematic. There will be feature requests, bug reports, training demands and a reliance on a developer\u2019s tacit knowledge.<\/p>\n

The solution<\/strong><\/p>\n

The AWS Instance Scheduler solution packages several AWS services together, including DynamoDB, Lambda and CloudWatch Events, into a deployable CloudFormation template, enabling users to automate custom start\/stop schedules for their EC2 and RDS instances. Instance Scheduler does not feature a user interface, with schedule configuration performed by using the DynamoDB interface or the
\ncommand-line interface.<\/p>\n

Instance Scheduler supports:
\n* Stopping and starting stand-alone EC2 instances
\n* Stopping and starting RDS instances
\n* Stopping and starting instances in multiple regions
\n* Starting or stopping instances manually outside the defined schedule
\n* Partial automation (stop-only or start-only schedule)<\/p>\n

It does not support:
\n* Stopping or starting instances that are part of an auto-scaling group (native
\nASG scheduled actions can be used in this case)
\n* Stopping or starting other managed services (e.g. Redshift, Elasticsearch,
\nElastiCache, etc.)<\/p>\n

 How it works<\/strong><\/p>\n

A DynamoDB table contains definitions of Periods and Schedules, with every Schedule having a unique name. Resource tags are used to associate EC2 and RDS instances with a particular Schedule (e.g. Schedule=uk-office-hours, when \u201cuk-office-hours\u201d is a name of one of the Schedules we defined in DynamoDB).<\/p>\n

So what are Periods? Periods contain conditions that define the hours\/days\/months an instance will run. Schedules are composed of one or more Periods.<\/p>\n

For example:<\/p>\n

Assume that instances need to run during office hours (Monday-Friday 9am-5pm ET) and also perform batch processing every night 2am-4am ET. To achieve this, we define 2 periods:<\/p>\n

\"aws<\/p>\n

The Period \u201coffice-hours\u201d covers our office hours requirement and is responsible for starting the instances at 9am and stopping them at 5pm Monday-Friday. The Period \u201cbatch\u201d starts the instances at 2am and stops them at 4am every day of the week. Now let\u2019s define our Schedule: <\/p>\n

\"aws<\/p>\n

\nThe Schedule \u201coffice-and-batch\u201d contains the two running Periods we defined (\u201coffice-hours\u201d and \u201cbatch\u201d) and specifies the time zone. With the Schedule defined, we can associate our EC2 and RDS instances with it by adding a resource tag with a tag key of \u201cSchedule\u201d and a tag value of \u201coffice-and-batch.\u201d Note that the tag value is identical to the name of the Schedule we defined. <\/p>\n

How much does Instance Scheduler cost?<\/strong><\/p>\n

The solution itself is free, however, because it utilizes a few AWS services under the hood (DynamoDB, Lambda, CloudWatch), and is deployed in your AWS account, you will still pay for these services. But, in our experience, unless you have a significantly large AWS environment, these costs are generally below $10 per month \u2014 and we\u2019ve yet to see a customer that doesn\u2019t save at least $10 in usage
\ncosts with Instance Scheduler. Essentially, it pays for itself.<\/p>\n

Instance Scheduler provides a reasonable set of features, at a minimal cost, and meets most AWS customer use cases. If you need to schedule resources other than EC2 or RDS, support multiple cloud platforms, or require a UI, however, Instance Scheduler might not be right for you.<\/p>\n

And of course, if you like the solution, but want someone to manage it for you, Rackspace can help. We provide full management of Instance Scheduler for any AWS customer with a managed service so you have a Fanatical Experience.<\/p>\n

As an added bonus, AWS Instance Scheduler improves security. A system that is only running for 30% of time has a lower security risk profile. Plus, it\u2019s good for the environment! You switch off the lights at home when not in use. Shouldn\u2019t you do the same with your infrastructure?<\/p>\n

If you\u2019re a current Rackspace customer and would like to start using Instance Scheduler, please submit a request via a Support ticket, or reach out to your account manager.<\/p>\n","protected":false},"excerpt":{"rendered":"

Dramatically Cut Costs with AWS Instance Scheduler<\/p>\n","protected":false},"author":2,"featured_media":8058,"menu_order":0,"comment_status":"open","ping_status":"open","template":"","format":"standard","meta":{"content-type":"","nid":"26592","content_type":"blog","created_by":"10130","authored_on":"2024-02-09T10:37:34+00:00","field_article_industries":"","tag_category":"Technical","field_blog_author_text":"by Maxim Rozin, Principal Engineer, Rackspace Technology","seo_description":"Oracle REST Data Services (ORDS) and SQL Developer. Develop RESTful web services and enhance database capabilities.\r\n","field_featured_image":"","insights_image":"\/wp-content\/uploads\/2026\/blog_technical_images\/26592-insights_image-aws-technical-blog-1-insights.png","field_partners":"","rxt_platforms":"Amazon Web Services (AWS)","product":"","publisheddate":"2019-02-25T10:37:34","rxt_solutions":"Public Cloud","opengraph_tag":"","x_tag":"","teaser_flag":"Technical Blog","seo_titile":"Dramatically Cut Costs with AWS Instance Scheduler | Rackspace Technology","field_topics":"","field_slug":"dramatically-cut-costs-with-aws-instance-scheduler","footnotes":""},"categories":[178],"tags":[204],"class_list":["post-8056","blogs","type-blogs","status-publish","format-standard","has-post-thumbnail","hentry","category-blog","tag-technical"],"_links":{"self":[{"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/blogs\/8056","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/blogs"}],"about":[{"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/types\/blogs"}],"author":[{"embeddable":true,"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/comments?post=8056"}],"version-history":[{"count":3,"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/blogs\/8056\/revisions"}],"predecessor-version":[{"id":9852,"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/blogs\/8056\/revisions\/9852"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/media\/8058"}],"wp:attachment":[{"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/media?parent=8056"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/categories?post=8056"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.rackspace.com\/en-gb\/wp-json\/wp\/v2\/tags?post=8056"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}