The service level agreement for Amazon EC2 includes several compute services, including: For EC2 services (and related services), the credits available are: In this case, we must multiply the availability target of each service. The following figure shows an example of a typical web application running on EC2. All S3 services have a guaranteed operating life of 99.9%, with the exception of the following services, which are guaranteed for 99% of operating time: ALS operating time varies by type of S3 service and guaranteed operating time ranges from 99.9% to 99%. AWS offers service credits for non-compliance with the ALS operating time, but it is important to note that they are not applied automatically. To obtain service credits for downtime, AWS customers must assert a right to credit by opening a business in the AWS assistance centre with the words “SLA credit requirement” in the subject line and specifying the details of the dates and times of defaults or failures for which you use secure credits with protocols and resource IDs for the services concerned. If service credits are available for loss, they are normally used as credit on future invoices for the same service. AWS support response times are separated from each SERVICE-Uptime-SLAs. Response times vary between the 3 AWS support plans available – developers, businesses and businesses. To facilitate the comparison, I have listed the available response times and associated support costs in the table below. Don`t expect a 15-minute support response if you only pay for a business support plan! Although, in my experience, AWS support is both fast and effective. This Amazon Compute Service Level Agreement (“SLA”) is a policy for the use of included services (below) and applies separately to each account that uses the services contained. In the event of a conflict between the terms and conditions of this ALS and the terms of the AWS customer agreement or any other agreement with us regarding your use of our services (the “Contract”), the terms of this ALS apply, but only to the extent of such a conflict. The terms used in this agreement, but which are not defined here, have the meaning of the agreement.

For all RDS areas hosted in multiple availability areas (the “Multi AZ” setting is defined as “True”), Amazon guarantees 99.5% operating time on each monthly billing cycle. This allows up to 3.65 hours of downtime per month. For applications that cannot tolerate downtime, customers should consider hosting their databases in multiple regions or using another database service, such as Amazon Aurora, which has an ALS with a 99.99% operating time. The examples presented give rise to an expected availability of 99.80% to 99.92% depending on the services concerned.