AWS Lambda Extends Timeout to 90 Minutes on Managed Instances, Six Times the Previous Limit

Overview

AWS has extended the maximum execution timeout for AWS Lambda functions running on Lambda Managed Instances to 90 minutes, up from the previous 15-minute limit, according to a September 9 post on the AWS Compute Blog. The new ceiling applies to asynchronous and event source mapping (ESM) invocations; synchronous invocations retain the existing 15-minute maximum timeout, according to AWS.

What We Know

• The new 90-minute cap is six times longer than the previous 15-minute limit, according to InfoQ.
• Lambda’s function timeout has grown over time, increasing from 5 minutes at launch in 2014 to 15 minutes in 2018, according to AWS.
• AWS says longer runtimes support workloads such as media processing, financial calculations, data processing and ETL pipelines, AI inference, and web scraping or large file transfers, according to AWS.
• Lambda Managed Instances now also support Graviton5-powered EC2 instances, a separate recent addition to the service, according to InfoQ.
• The extended timeout is available in every region where Lambda Managed Instances are offered, according to InfoQ.

Idempotency and Retry Risk

Longer-running functions widen the window in which retries can produce duplicate work. AWS cautions that “Lambda does not guarantee exactly-once processing. With longer-running functions, the window for retries and duplicate deliveries increases,” according to AWS. For asynchronous invocations that fail or time out, “Lambda applies your configured retry policy (up to two retries by default),” according to AWS.

Community Reaction

Reaction to the change has been mixed. Yan Cui, described by InfoQ as a “serverless expert and AWS Hero,” wrote that “Part of me is sad that this further blurs the line between ‘running a server’ and a Lambda invocation. But it does make sense, especially for the ‘Lambda for agentic workflows’ use case,” according to InfoQ.

Rajesh Pandey, a principal engineer at AWS, said, “I’ve lost count of how many times customers have asked when are you guys going to build, ‘Long-running Lambda functions.’ Finally, it’s here. A lot of the duct tape built around Lambda’s maximum execution duration can now come off,” according to InfoQ.

Not every reaction was positive. A user identified by InfoQ as Dull_Caterpillar_642 warned that “If it’s taking 90 min, there’s a good chance you should be using something like durable functions (…) Unless your code is literally doing meaningful stuff that entire time as opposed to waiting on other stuff, in which case the economics of Lambda may not work out as much in your favor, and something like ECS Tasks or AWS Batch could make more sense,” according to InfoQ.

What We Don’t Know

AWS has not said whether the 90-minute cap will eventually extend to synchronous invocations or whether further increases are planned. The announcement did not address pricing implications for functions that run substantially longer than before.