AWS users usually assume that Northern Virginia, also referred to as US East (N. Virginia) and us-east-1, is the least reliable in terms of uptime.
We analyzed AWS outage history in 2022 across regions to see if N. Virginia, indeed, had the most downtime.
Then we reviewed and proved some of the theories as to why N. Virginia has the most outages.
The Data Behind the Study
We used our product, StatusGator, as the data source. StatusGator aggregates data from official status pages for an all-in-one team status page. Our unique dataset allowed us to understand the reliability of AWS Regions between January 1 and December 31, 2022. Of course, the official status page will only reflect widespread AWS outages, but using that data we can understand which regions have the most major incidents.
We excluded both GovCloud Regions: US-East and US-West, since these are not available to the general public.
Ranking of the AWS Regions by the Number of Outages and Duration in 2022
|Regions||Number of outages||Duration|
N. Virginia ranked No. 1 as the least reliable Region of AWS, with a total of 23 partial outages in 2022, lasting 61 hours and 7 minutes.
It looks like not much has changed since 2021 and N. Virginia remains the least reliable region.
Why Is Northern Virginia the Least Reliable AWS Region?
We took the three most popular assumptions among Reddit users and checked the data to find out why Northern Virginia is the least reliable AWS Region.
Assumption 1: Northern Virginia Region Has More Services
The assumption is that the Northern Virginia region has more services, and consequently, more things can go wrong.
Let’s take a look at the number of services offered by the regions and the number of outages in those regions.
|Regions||Number of services||Number of outages|
We observed that, indeed, N. Virginia offers the biggest number of services – 215. The next largest regions are Oregon with 207 services, and Ireland with 204.
The difference in the available services across these three largest regions is not that big. So, it appears unrealistic that offering only a few services more is the cause of N. Virginia’s downtime.
Looking at the top five largest AWS Regions, we can’t see much of a correlation between the number of services and outages.
If the assumption were true, Oregon and Ireland, both would have had more outages. But Oregon had 5 outages, while Ireland only had 1 outage in 2022. To prove the point further – Sydney and Tokyo, the No.5 and No.6 largest regions did not have any outages in 2022.
Additionally, we wanted to look at the most popular services as a cause of disruptions. So, Amazon Elastic Compute Cloud (EC2) is the most popular AWS service. Here’s the outage history of EC2 across the regions in 2022:
EC2 Service Outages and Duration across the Regions
|AWS Region||Number of outages||Duration of outages|
We can observe that EC2 outages in N. Virginia cover 50% of the total outage duration for this service.
Ohio had the same number of EC2 outages – 4 incidents. But the duration is nearly 3x lower compared to N. Virginia.
It leads us to believe that the number of services is not responsible for the reliability of the Northern Virginia Region.
Based on our findings, Assumption 1 is probably not true.
The number of services available in N. Virginia is not much bigger compared to other AWS Regions to explain the difference in outages.
Assumption 2: Northern Virginia is the Largest and Most Used Region
Some Reddit users believe N. Virginia is the most popular region and is heavily loaded with customers running numerous operations simultaneously.
To verify, we looked at the number of StatusGator users who monitor the status of different AWS Regions.
According to our data, N. Virginia is the most monitored AWS Region and likely has more users than other regions.
The number of users who monitor N. Virginia is 189% more compared to the second-largest region – Oregon, and 214% more compared to the third-largest region – Ohio.
Assumption 2 is most likely to be true.
Due to a higher number of users watching the status of N. Virginia, we assume that the Region has more customers using its services at the same time causing an overload.
Assumption 3: Northern Virginia Is the First Region, and Was Set Up Differently
Some users believe that N. Virginia, being the first-ever AWS Region could be set up differently.
We have doubts that N. Virginia has a different infrastructure causing it to be less reliable, and AWS does nothing to fix it seeing that other setups work better.
We could speculate that due to the biggest number of users, they are not risking major changes. But in this case – the data is not in the open to prove Assumption 3.
We can not verify Assumption 3, but we assume it’s probably not true.
If it was true, then why hasn’t AWS fixed it?
We confirmed that us-east-1 (N. Virginia) remains the least reliable AWS Region in 2022.
We saw N. Virginia contributed 51.11% of the total number of outages across the AWS Regions.
As for duration, 68% share of the outage duration belonged to the Northern Virginia Region.
We saw that more users watch the status of N. Virginia. Specifically, 189% more compared to the Oregon Region, and 214% more – compared to Ohio.
Based on this, it is likely that N. Virginia is catering to the highest number of users and has the largest capacity.
To monitor the status of Amazon Web Services, its components, or regions – try StatusGator for free.