Tips and tricks for logging and monitoring AWS Lambda functions
Yan Cui
925

Thanks for the article. IOpipe will have some solutions to help with managing your account concurrency limit very shortly. This will be done first, by supporting project, account, and region level alerts… and then through increased visibility into concurrency under the covers by exposing process, container, and VM lifecycles.

While security issues with npm are out of scope for us, we’re proud to say that we do not have any dependencies in the base IOpipe NodeJS library, simplifying the security concerns. Our library is developed and maintained in the open and kept intentionally lightweight for reasons of performance, reliability, maintenance, and for simplicity of 3rd-party audits.

The points on user-facing latency on API Gateway (and Alexa) sourced events are valid, if the additional 10–30ms is impactful to your application. I find that many users use Lambda for non-interactive events, where we have no effective penalty as connection to our service follows function execution. However, a large focus of our 1.0 release was to minimize the latency impact through a variety of tricks around TLS Session Reuse, HTTP keepalive, asynchronous DNS queries and smart caching, etc. We have a number of additional improvements we are seeking to make, to improve performance further.

We’d be pleased to discuss this further in private and see how IOpipe can meet your needs, now and in the future.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.