Jun 10, 2024
By Russ Savage
Jun 10, 2024
By Russ Savage
Earlier this year, we opened up a developer preview of our new cloud based traffic inspection tool and since then, we’ve been busy responding to customer feedback and adding some cool new capabilities.
Today, we are excited to announce that we have rolled out live updating, additional retention days, replay requests to your backend with modifications, and filtering by specific time range. With these changes, we are officially marking our traffic inspector as generally available and ready for your production workloads. By consolidating observability into a single location in the ngrok dashboard, we’ve made it faster and easier to debug your applications.
At this point, we recommend anyone who is still using the ngrok agent local inspect interface to try the new traffic inspector in your ngrok dashboard.
Over the past year, we’ve taken our standalone ngrok agent and decomposed it into idiomatic open source libraries that can be embedded directly into your applications. We’ve seen users build some pretty amazing things with them, but one of the biggest challenges they’ve faced is not being able to see the traffic that was going through their endpoints. This is a key feature of the ngrok standalone agent, but something that didn’t quite make sense as an embedded library.
With the release of our new cloud based Traffic Inspector, it doesn’t matter what agent you use to connect to ngrok, all of the traffic shows up in a single place in your ngrok dashboard. You can also replay traffic to those SDKs the same way you would the local standalone agent, which is a powerful debugging tool.
When we started building our cloud based traffic inspector, we wanted to build an experience that was better than the local traffic inspection interface used by thousands of ngrok users every day. We knew that one of the most difficult experiences to replicate in the cloud would be getting traffic to show up in near real time in the browser.
Today, you can choose between a live view that is useful for debugging your application in real time or a static view that can be helpful when troubleshooting a production issue. When you select a specific date range, the live updates automatically pause so you don’t lose your place.
When we rolled out the ability to capture and view traffic to your HTTP endpoints, we started with a small lookback window of three days on purpose. Building a system to capture the sheer volume of traffic we process on a per minute basis comes with a lot of challenges, and we didn’t know how this experience would perform for users with any significant amount of traffic.
Throughout the past few months, we were able to analyze and optimize the data storage and queries to a point where we felt comfortable extending the retention period to up to 90 days. Users of our production plans can add this to their existing plans or purchase is directly through the app.
Also, now that you have an increased retention window, we wanted to make sure that you could get to the traffic that you want to look at. We added the ability to filter based on a specific time range for the requests so that you can find what you are looking for quickly. This is very useful when looking at an error our outage that happened at a specific time.
When building out a new application that responds to external traffic like a webhook or callback, it’s incredibly helpful to be able to capture and replay requests to your service without having to manually go back and trigger them. This could be very cumbersome: an action has to be simulated in a 3rd party system, which in turn triggers the webhook, or physically calling a phone in the case of services like Twilio.
We are excited to announce that you can now customize and change the request before replaying it to your upstream service. This allows you to swap out headers, update credentials, or change values in the body to see how your service behaves.
In addition, while triaging production outages, you can replay the exact request that caused the original error to occur, but send it to a different endpoint, which contains the updated code. This allows you to move faster without having to wait for the anomaly to happen again.
All of this is available right from our dashboard. Troubleshooting doesn’t have to be a burden - with Traffic Inspector, it’s easier, faster and even fun! See the power of Traffic Inspector for yourself.
We are also very grateful to the thousands of users who have been using our developer preview of traffic inspector and providing feedback. Thank you. Without your input, we can’t build an amazing product. We’ve also enjoyed reading about how much our users love the power of our platform.
We are excited to look for more opportunities to provide insights into the traffic running through your ngrok account. We have plans for advanced filtering and more tightly integrating this tool into the cloud edges and endpoints that are being created.
Give it a spin and let us know what you think. We’d love to hear from you in our new ngrok Community Repo.