- 最新
- 投票最多
- 评论最多
From the user perspective, load tests tools like JMeter, Gatling and DLT can give latency metrics.
CloudWatch metrics for CloudFront and ELB also provides good insights regarding latency, but for Inter-Services it's probably best to have a APM (Dynatrace, AppDynamics) and/or a tracing solution, like X-Ray or Zipkin.
Hey there AWS-User-0484764,
I completely understand your concern about latency issues when it comes to a distributed application running on AWS. It's a common headache, especially when catering to B2B users across regions. Kudos to you for proactively addressing this!
From my own experience, JMeter, soapUI, and CloudWatch/X-Ray are indeed solid choices. However, if you're looking for a more streamlined and comprehensive solution, you might want to consider using Apache Kafka along with Apache Kafka Streams for real-time monitoring and analysis of application latency.
Kafka is renowned for its scalability and low-latency data streaming capabilities. By instrumenting your application to produce latency data to Kafka topics, you can then employ Kafka Streams to process and analyze this data in real-time. This allows you to gain deeper insights into not just network and inter-AWS service latency but also your application's latency.
Furthermore, consider integrating AWS Lambda into the mix for auto-remediation actions based on predefined thresholds. This combination should give you a robust solution to monitor and address latency concerns efficiently.
相关内容
AWS 官方已更新 3 年前
