There is an ongoing, continuous flow of existing and new workloads moving to the cloud. One significant side effect of this transition to the cloud is that many of these applications, systems, and workloads are becoming critical to the teams using them. It may not be immediately apparent that this is, in fact, the case, but it is a crucial aspect that teams need to factor into their overall project and deployment plans.
Ensuring that an application or a workload is highly available is essentially a two-stage process – at least, this is how I see the challenge. Firstly, there is the level of availability of the underlying cloud services, and secondly, there is the application framework’s capability to support an outage of the database. Most cloud vendors provide service-level agreements, but these agreements come with all sorts of caveats. Those caveats make it difficult for IT and business teams to deliver a highly available application that meets specific user expectations and business objectives.
With Autonomous Database, Oracle provides a service level agreement of 99.995% availability where Autonomous Data Guard is enabled and 99.95% availability where it is not enabled. Firstly, what do those two numbers mean? Where Autonomous Database is enabled for your Autonomous Database, Oracle guarantees a maximum of around 2.10 minutes of downtime per month. If you don’t enable Autonomous Data Guard, Oracle ensures only 21.36 minutes of downtime per month. There is a lot more information about the service level agreement for your Autonomous Database in this blog post by my colleague Nilay Panchal – Autonomous Database now provides a 99.995% availability SLA with Autonomous Data Guard.
The other aspect of “availability” worth considering is the overall availability of the data center where your Autonomous Database is located. How reliable is that region? How does one region compare to another region in the same geographical zone? Having that sort of information could be helpful to IT and development teams. Therefore, we published a new web page on oracle.com to help you understand the overall availability metrics for Autonomous Database across our various regions. The metrics table covers a rolling three-month period and is updated daily. Please consider this a regional-level view of the same availability data we provide on your Autonomous Database console. See this section “Monitor Autonomous Database availability” in the documentation.
The new regional availability data page on oracle.com looks like this:
What do these numbers mean, and how do we collect the data shown on this page? The availability data shows a list of average availability percentages across our data centers in each region. These availability numbers cover a whole month, including the maintenance windows, with no exclusions. The aim of providing this level of detail is to give you a real-world example of the general level of availability that you should expect to see for your Autonomous Databases. It’s sort of a health check report for each of our regions.
I am unaware of any other cloud vendor providing this level of information for their database services. Essentially, we are being as transparent as possible about how our services perform so you can be confident about the level of availability for your mission-critical applications and workloads running on the Autonomous Database.
We hope you find this page useful, and if you have any suggestions about how we can improve the way the data is presented to make it easier to understand, please let us know.
Here are some other blogs posts that will help you explore the general aspects of availability in a little more detail:
- Announcing Autonomous Data Guard!
- Cross-Region Autonomous Data Guard – Your complete Autonomous Database disaster recovery solution!