Reduce immediate hourly commitment of AWS RIs in exigencies
A tip to spread your No-upfront, Convertible RI commitments over a longer duration and reduce immediate $/hr spend.
Pravanjan Choudhury
October 18, 2021
table of contents
SHARE
If you run AWS deployment, you would know that Reserved instances (RIs) are a great way to obtain a discount on your always running workloads (base compute) by committing for a certain duration like 1 year or 3 years.
Typically, you would look at your base on-demand compute to determine how many RIs to purchase at any time. This is all great in the normal course i.e. your business is growing and you are increasing the cloud spend. Also possibly buying more RIs from time-to-time.
Now, consider an exigency like a large customer churn, re-architecture, or an incident like covid onset that reduces your on-demand compute base. In these cases, your on-demand RIs may not get fully utilized temporarily or you may opt to run the machines anyway at reduced utilization!
Here is a tip to spread your No-upfront, Convertible RI commitments over a longer duration and reduce immediate $/hr spend for temporary relief.
The example shows 3 RI Line-items purchased sometime in the past, amounting to 26 m5.2xlarge machines, expiring on June 30th, 2022. Assume Today to be 15th Oct 2021. You can simply reserve one t3.nano (the smallest instance) today for 3 years (no upfront, convertible), thus expiring on 15th Oct 2024. When you select all 4 RIs Line-items and put them up for exchange, the number of machines reduces to 6 from 26 with an expiration date of 15th Oct 2024.
The way RI exchanges work is that AWS takes the SUM $ commitment value of the RIs to be exchanged and spreads it over the Max expiration date of the RIs to be exchanged. What we have essentially done is spread the 6 months of leftover commitment over 3 years. Read AWS docs here.
Few points to note here
This process is irreversible, i.e. say your usage is back, you can't move back the RIs to the original state as the exchange date always moves forward. So you may have to purchase new RIs.
You have a longer commitment with smaller $/hour than earlier. The total commitment remains the same
Applies to No-upfront, convertible RIs only and not for other mechanism like AWS Savings plan.
You can of course sell RIs but some restrictions apply like you can't sell Convertible RIs at the moment. AWS docs here.
Stay tuned for more tips on how to choose between different reservation instruments like Standard, Convertible and Savings Plan wisely.
"With Facets, our staging environments look identical to production environments. So in case of production issues, we can be sure there are no infra drifts."
Kadam Jeet Jain, Co-Founder & CTO, Treebo Hotels and Hotel Superhero.
GGX switched from AWS to GCP in 2 weeks
"Facets has radically changed our DevOps for the better. They did all the heavy lifting and saved us precious time and resources in our when we switched from AWS to GCP."
Kaustubh Bhoyar, Head of engineering, GGX
Trusted by companies to run production at scale
Request a Quote
Let us know if you have any additional queries, we'll get back to you soon.
Capillary reduced ops tickets by 95%
“Our releases are fast. And with less developer time needed our teams can focus on building exciting features. We’ve saved countless hours and costs.”
"With Facets, our staging environments look identical to production environments. So in case of production issues, we can be sure there are no infra drifts."
Kadam Jeet Jain, Co-Founder & CTO, Treebo Hotels and Hotel Superhero.
GGX switched from AWS to GCP in 2 weeks
"Facets has radically changed our DevOps for the better. They did all the heavy lifting and saved us precious time and resources in our when we switched from AWS to GCP."
Kaustubh Bhoyar, Head of engineering, GGX
Trusted by companies to run production at scale
Get in touch with us
Tell us your queries and we’ll get back to you
Prefer email? Reach out to us at info@facets.cloud