This course has been updated! We now recommend you take the AWS for Front-End Engineers (ft. S3, Cloudfront & Route 53) course.
Transcript from the "Locust & Rollbar" Lesson
>> Kevin Whinnery: Now, I'm gonna talk a little bit about a few tools that we found really useful in monitoring and sort of preparing our applications for production. The two aspects that I'm gonna focus on and give you a little demo of this afternoon are a load testing tool called Locust.
[00:00:23] And a error tracking service that lots of teams use at Twilio actually called Rollbar, which we'll take exceptions in our application. And put them in a service that will notify us if bad things happen in our application wall it's running in production. So I'll do a quick demo of locust, and then I'll turn you all loose on starting to add in some error handling in our application on the server side, which we haven't actually done yet, if you have might have noticed.
[00:00:55] So the first thing I'm gonna show you is a tool called Locust on their load testing tools for just about every platform. This particular one is a Python based tool that I have favored because it's relatively easy to use. It gives really good reporting output and is configurable through a web interface, which I'll show you here in a second.
[00:01:20] It's got a large community around it. And it allows you using Python code to actually simulate different types of usage scenarios on your website. So the flexibility of the configuration and the maturity of the tool as well as sort of the general usability are some reasons why I favor it over other things that are out there.
[00:01:43] The production monitoring tool, we're gonna take a look at is called Rollbar. It's a third party service in whose API you integrate into your application. And when handled exceptions are raised, you can send that exception data to this service. And it'll send out notifications to whatever provider you happen to have configured.
[00:02:08] At Twilio, we use pager duty to create a schedule of folks that are responsible for responding to issues and outages on the website.
>> Speaker 3: So, Rollbar does server side stuff too?
>> Kevin Whinnery: It's fully a server side, it's not quite. Yeah, I think you might be able to do client side, but we only use it for the server side.
>> Speaker 3: You can use track to ask in the clients side.
>> Kevin Whinnery: I know you can use track to ask on the client side.
>> Audience: [LAUGH]
>> Kevin Whinnery: Minor plug for our hometown hero, Todd Gartner.
>> Speaker 3: Yeah.
>> Kevin Whinnery: But yeah, so Rollbar is used across many services at Twilio to notify folks that things are going wrong, and it's something that I've really come to enjoy.
[00:02:52] If you have identifying information about a user at the time an exception is thrown. You can, in your Rollbar output, see which user an exception was generated from the origin. If you have Rollbar configured correctly, you can get a lot of very useful information about any exceptions in your application, so definitely worth checking out.
[00:03:16] And it integrates very nicely with Express applications as you will find out when you check out the exercise.