- Newest
- Most votes
- Most comments
DynamoDB tables do not spin down/turn off. With DynamoDB, you are not setting up instances, or managing servers, its completely serverless (think of S3 and Lambda). With on-demand mode, you pay per request, so if you don't make any requests you don't pay for anything (except the data storage).
On-demand tables will instantly accommodate a spike in traffic up to a minimum of your warm-throughput values. For example, lets imagine your warm throughput values are 4000 WCU and 12000 RCU, that means that even after days/months/years of no usage, you can instantly scale to 4k WCU and 12k RCU.
An On-Demand AWS DynamoDB table does not automatically stop or shut down after a period of inactivity. DynamoDB on-demand mode is designed to be truly serverless and scales automatically to accommodate workloads without capacity planning. It can scale down to zero when no requests are being issued against the table, but the table itself remains active and ready to serve requests at any time.
The on-demand mode allows DynamoDB to instantly scale up or down to meet demand, ensuring that your application can respond to requests without any downtime or need for manual intervention. This means that even if your table experiences long periods of inactivity, it will still be available and ready to handle requests as soon as they come in.
It's important to note that with on-demand mode, you only pay for the actual read and write requests you perform, so during periods of inactivity, you're not charged for unused capacity. This pay-per-request pricing model makes on-demand mode cost-effective for applications with unpredictable or variable traffic patterns.
The only way to actually stop or shut down a DynamoDB table is to explicitly delete it using the DeleteTable operation. When you delete a table, it and all its items are removed, and any associated streams or indexes are also deleted. However, this is a manual operation and not something that happens automatically due to inactivity.
Sources
DynamoDB on-demand capacity mode - Amazon DynamoDB
Introducing configurable maximum throughput for Amazon DynamoDB on-demand | AWS Database Blog
DeleteTable - Amazon DynamoDB
@Riku_Kobayashi @re:Post Agent Since On-Demand DynamoDB does not stop automatically, what is the minimum capacity (memory, instances, or resources) that remains active? After a period of no usage, how long does it take for DynamoDB to scale back to its minimum level?
Relevant content
- asked 3 years ago
- asked 2 years ago
- asked 9 months ago
- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated 2 months ago
I have a few questions about how DynamoDB scales beyond this:
If my workload suddenly exceeds 4,000 WCU and 12,000 RCU, how long does it take for DynamoDB to scale up? Is it truly instant, or is there a delay? Once scaled up, when does DynamoDB scale back down to the original warm throughput levels? Is there a cooldown period, and if so, how long does it take?
If you go above 4k/12k then you will be throttled initially, and it can take several minutes to scale up. However, if you anticipate going above that level, you can pre-warm your warm throughput values to any level you like. Warm throughput values never decrease.