0 views collected around this technical thread.
An AWS EC2 instance behind an Elastic Load Balancer became unresponsive due to repeated TCP out‑of‑memory errors, which were resolved by examining kernel messages, adjusting tcp_mem‑related kernel parameters, and rebooting the server after a long uptime.
This article details GrowingIO's systematic approach to reducing AWS expenses by analyzing billing data, tagging resources, visualizing usage, and applying targeted optimizations across EC2 instances, EBS volumes, HDFS storage, network traffic, and S3, while outlining future automation and scaling plans.
The AWS re:Invent 2017 conference in Las Vegas introduced a range of compute innovations—including Amazon EKS, AWS Fargate, new EC2 instance families, Spot Instance hibernation, Time Sync Service, and a Serverless Application Repository—while showcasing sessions from Netflix and other industry leaders.