Situation
As yesterday December 7th a new AWS outage made headlines, it seems opportune to talk about some AWS Services that are global (and not only tied to one region endpoint) like IAM and S3. For some Cloud Operations teams, to face an AWS outage in some regions is to have blockers on their daily jobs. The purpose of this post is to show how to continue doing admin jobs during a partial outage.
Issue
AWS reported issues with their APIs on North Virginia (us-east-1) region. And let’s suppose we are on a Cloud Operations team and we need to perform some tasks in AWS Console, like for example, create a S3 Bucket and create a new user that have access to the given bucket (all in region US East 1)… How will be supposed to perform those actions with AWS issue without losing all day?
Steps
Login into the AWS Console with your user and password. If you are redirected by us-east-1 region, go to your browser url and change it with:
https://<region>.console.aws.amazon.com/
where<region>
is a valid AWS region (for example, in this case we’ll use the Ohious-east-2
region).
So now we are logged into the AWS console successfully! We can then go to
Services
->IAM
.
Clicking onUsers
->Add users
we can create a new user and assign theAmazonS3FullAccess
AWS Managed Policy (just an example, it can be any other permission).
After creating our user, let’s create a new S3 Bucket from where we can read/write data (keeping logged in a different region). So we can go to
Services
->S3
.
Then we create a new bucket. Provide a unique name for that S3 Bucket and then on the region select theUS East 1
. Then, go to the end of the page and click onCreate
.
Conclusion
Voilà! We did our task but located on a different region! Although it is better to have AWS working 100% on all the regions, and that this approach cannot be applied on several services. This can be a temporal workaround to avoid getting stuck when performing some Cloud Operations tasks.
Top comments (0)