Question about how the instance fee is applied

Question

In a Windows On-Demand fleet type, let's say that the instance fee is $0.10 per hour.  According to the pricing page (aws.amazon.com/appstream2/pricing), "Windows On-Demand fleet instances are charged hourly."  Does that mean if a user logs into an instance and stays for 5 minutes and then logs out that this incurs a $0.10 charge (the charge for a full hour), not $0.008 (the charge for 5 minutes)?  And if the user logs in again shortly after logging out, are they charged $0.10 AGAIN?

I ask because that's what my billing statement for last month makes it look like.  It says "$0.1 per hour for running a stream.standard.medium Fleet instance in AWS region US East (N. Virginia)," and it says quantity 16 hours.  I was the only one who logged in last month, and my total session time was probably less than ONE hour.  I may have logged in 16 times, but not for a total of 16 hours.

Given that I currently have set my user login sessions to have a limit of 15 minutes per session, if what I'm asking is true, should I set the session limit to, say, 55 minutes so that the instance stays running for most of the hour, since I'm paying for it anyway?  Or is there a way to have the system only charge me for actual logged-in time, not rounded up to the next hour?  Thanks.

PS:  I got an automated "AI" response to this question, which suggested that I could consider using certain AWS instance types that have per-second billing.  Are there such instance types for Windows in the General-Purpose category?  Or per-minute?

re:Post Agent · Answer

Based on the information provided in the AWS EC2 pricing documentation, it appears that your understanding of the billing for Windows On-Demand instances is correct, and this explains the charges you're seeing on your billing statement.

For EC2 On-Demand instances, which includes AppStream 2.0 fleet instances, you are charged for each hour or partial hour that an instance is running. This means that even if you only use the instance for 5 minutes, you will be charged for a full hour. If you log out and then log back in within the same hour, you won't be charged again for that hour. However, if you log in again in a new hour, that will start a new billing period.

Given your usage pattern and the current 15-minute session limit, it's likely that each login is triggering a new hourly charge, even if the actual usage time is much shorter. This explains why you're seeing 16 hours charged when your total active usage time was less than one hour.

Your suggestion to increase the session limit to 55 minutes is a good one. This would allow you to maximize the value of each hourly charge by keeping the instance running for most of the hour you're paying for anyway. This way, if you need to log back in within that hour, you won't trigger a new hourly charge.

Unfortunately, there isn't a way to have the system charge only for the actual logged-in time at a granularity finer than one hour for On-Demand instances. If you need more precise billing based on actual usage time, you might want to explore other pricing options that AWS offers, such as per-second billing for certain instance types or considering reserved instances if you have predictable, long-term usage patterns.

Remember to regularly review your usage patterns and adjust your configuration to optimize costs while meeting your needs.  
**Sources**  
[Amazon EC2 Dedicated Instances Pricing | AWS](https://aws.amazon.com/ec2/pricing/dedicated-instances/)  
[EC2 On-Demand Instance Pricing – Amazon Web Services](https://aws.amazon.com/ec2/pricing/on-demand/)  
[Amazon EC2 - How AWS Pricing Works](https://docs.aws.amazon.com/whitepapers/latest/how-aws-pricing-works/amazon-ec2.html)

Question about how the instance fee is applied

Add your answer

Relevant content