Fargate Service Performance Degedation
On 04/21 at about 11:23, we started experiencing a severe degradation of performance on our Fargate deployed services. Tasks that were taking milliseconds are now taking seconds. We did not deploy any new services, these are the same services that have been running for months.
The only interaction these tasks have is with an RDS Aurora databases.
Has something changed on AWS that I need to account for?
Thank you,
Chris
- Newest
- Most votes
- Most comments
Look for the cause of poor performance.
Are you seeing anything unusual in the application logs, for example?
If you are using RDS Aurora, you can also enable "Performance Insights" to search for SQL that is taking a long time to process.
https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/USER_PerfInsights.Enabling.html
You can also check AWS Compute Optimizer and other tools to see if ECS is underperforming.
https://docs.aws.amazon.com/compute-optimizer/latest/ug/view-ecs-recommendations.html
Relevant content
- asked 4 years ago
- asked 15 days ago
- asked 7 months ago
- asked 2 years ago
AWS OFFICIALUpdated 3 years ago- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated 10 months ago
In addition to the application logs and RDS Aurora performance insights, I would recommend looking at the CloudWatch logs for either ECS and EKS (unclear which service is in use, Fargate is available for both). These will provide visibility into issues with the launching of new containers, or issues with the container environment.