Dr. Werner Vogels 2020 re:Invent keynote
Often, the announcements made during the keynote by Werner Vogels are mostly aimed at developers. This trend has continued this year but I believe that the announced services are also very useful to other figures who work in the cloud field.
Let’s go see them:
AWS CloudShell environment is based on Amazon Linux 2 and offer 1 GB of persisting storage in the $HOME directory per region at no additional cost. Any data stored outside of this directory will be deleted once the shell is closed. This means we can test new applications by installing them outside the $HOME directory and if something goes wrong we can just start a new shell.
AWS Fault Injection Simulator (coming in 2021)
A famous phrase from Vogels says: Everything fails, all the time.
So it’s not a question of IF, but of WHEN.
AWS Fault Injection Simulator is a service that create chaos in an application in order to easily discover the weaknesses and performance bottlenecks giving the engineers the possibility to intervene to solve them by improving performance, observability and resilience, before this happens maybe in production putting sensitive data at risk. It’s a fully managed service so it is very easy to start testing without the need to install any agents. Using rules based on Amazon CloudWatch alarms, tests can be stopped automatically if a service drops below the set threshold. At the end of a test, you will be able to see what actions have been performed, whether your environment has behaved as expected and much more.
“Logs are a source of truth” (cit. Werner Vogels)
Grafana is a cross-platform open source web application for interactive analysis and visualization capable of centralizing data from multiple open-source sources such as Prometheus, Icinga, ElasticSearch and Splunk.
With Amazon Managed Service for Grafana you can analyze, monitor, and alarm on metrics and logs using a single dashboard for all the data sources and being a managed service it automatically manages the processing infrastructure. It’s integrated with all AWS services, allows you to create a workspace in one account and access data sources in all your accounts and regions, thus centralizing the monitoring and logging system and also integrates with AWS Single Sign-On ensuring access to dashboards only for authorized users. Still in preview, the service is available in US East (N. Virginia) and Europe (Ireland) regions.
Prometheus is an open-source metrics-based monitoring system based on multi-dimensional data model, scalable data collection, and a powerful query language.
Amazon Managed Service for Prometheus is a monitoring service compatible with Prometheus that simplifies the monitoring of applications running on containers, integrated with EKS, ECS, Amazon Managed Service for Grafana and AWS Distro for OpenTelemetry and use the open source Prometheus query language (PromQL). Being a managed service, it automatically adapts to importing, querying and storing data according to the workload. Offers multi-AZ replication and use IAM to authenticate and assign fine-grained permissions to users and groups.
AWS Distro for OpenTelemetry
OpenTelemetry is a collection of tools, APIs, and SDKs which you can use to instrument, generate, collect and export metrics, logs and traces for analysis in order to understand the performance and behavior of your software.
AWS Distro for OpenTelemetry is a secure, production-ready, AWS-supported distribution of the OpenTelemetry project. With AWS Distro for OpenTelemetry, you can instrument your applications in one go to send related metrics and traces to AWS monitoring solutions such as Amazon Cloudwatch, or AWS X-Ray and to third-party monitoring solutions such as Grafana or Datadog and use auto-instrumentation agents to collect traces without changing the code. AWS Distro for OpenTelemetry can be used to instrument application running on EKS, ECS and Fargate as well as on-premises. AWS Distro for OpenTelemetry for container services, and EC2 can be deployed and configured by using AWS CloudFormation templates, AWS CLI, or by using the Kubectl commands.
This should have been the closing keynote of this unusual 2020 re:Invent but which AWS wanted to extend to January 2021. A somewhat strange appointment for those who are used to attending in person in Las Vegas but who nevertheless saw aside AWS to release many services and features that will help build, maintain, and monitor better solutions.
And remember some of Werner’s mantras, “Encrypt everything” and “Log everything”