Sentry’s Post

Transcript

Hi everyone, I'm Rachel, Senior Product Manager working on Today, we're showcasing several major updates aimed at simplifying several major updates aimed at simplifying the way Sentry helps you triage and fix critical issues. While Sentry is known for surfacing problems quickly, over time our user experience has become a bit cluttered. We know that developers are under pressure to ship code fast, and we want to make sure that Sentry doesn't add friction to that process Our latest releases address these pain points by delivering more targeted and actionable insights. First, we'll touch on recent additions to the Core Issues experience, followed by what's new with uptime monitoring, and we'll wrap up with an announcement from the Session Replay team. To kick things off, I'll share an update on two of our most requested features, Better Search and Issue Views. Sentry Search has always been powerful, though with specific search syntax can be hard for new users to pick up. We've upgraded the whole experience to make searching for key terms more accurate with improved autocomplete suggestions and performance enhancements. The new search has been extended to the rest of Sentry, creating a consistent experience no matter if you're an issues, traces or replays. Now that it's easier than ever to search for your issues, it's only natural that we would make it easier for you to save and revisit key searches, simplifying future investigations. Whether you're monitoring specific problems or debugging across multiple projects, you don't want to recreate the same queries every time. With views, you can now easily save key searches and access them with a single click. This allows for more flexibility and personalization. With these updates, we paved the way for more effective debugging, which naturally complements our new uptime monitoring initiatives. In today's world, any downtime could have a huge impact on user experience and business operations. With uptime monitoring, you can keep an eye on your most critical endpoints and get instant notifications when services go down. Or experience degraded performance. Uptime can be configured as a new alert for a specific URL and it will be continuously monitored for any downtime failures through consistent Http requests. As soon as a downtime is detected, a new uptime issue will automatically be created with traces and related errors. In this example of an uptime issue, my configured URL returned a 500 error and Sentry traces back to a back end error revealing that my Next.js app was unable to connect to the database. This level of insight helps me quickly pinpoint the root cause of the downtime. Our goal with uptime monitoring is to give you full visibility into your system's health so you can make sure that no issues, from DNS resolution problems to application errors, fall through the cracks. Uptime Monitoring is an open beta, which means you can still try it out for free. Go to the Alerts tab in your Sentry org to get started. While catching bugs and maintaining a smooth workflow is crucial, understanding how users interact with their app is just as important, especially when things go wrong. I'm going to pass it over to Jasmin to share updates on how we're making that easier for mobile developers. Session replay for mobile is now in beta, with plans to officially launch at the beginning of next year. This means you have free, unlimited access to replays while it's in beta. We now support React Native and Flutter, as well as native iOS and Android applications including Swift UI and Jetpack Compose. Reproducing bugs can be tough, but with replay you can see exactly what led to an error, including crashes and debugging contexts like network calls, battery life, and connectivity all captured from the user session. So you can reproduce and fix issues faster. When building mobile replay we had two key focuses. One, protect user privacy with strict out-of-the-box privacy controls, and two, make it as easy as possible. For any developer, no matter their stack to get up and running quickly, let's look at it in action. This is my demo. A Empower Plant, an ecommerce app where you can buy plants and different supplies for them. I just got alerted to a new issue via Slack. It shows that I have a slow database query. Once I click in, I can see that this issue is impacting thousands of users. On one of my critical screens, the product screen, I can also see the duplicate database requests. While this doesn't look great, I want to truly understand the user impact of this performance issue. So I'm going to watch the replay, which is conveniently attached to the issue. I'm brought to the timestamp of the same issue. Playing the video, I can see exactly what the user experienced. And they're likely to leave since the product screen isn't loading quickly. The product screen finally loaded. Notice that out-of-the-box, all information, including text, images, and vectors is redacted. You can use replay without being worried of leaking sensitive user information. On the right side, I also see user interactions like taps and navigations. Let's Scroll down. I see another Flask error here in addition to the N + 1 query issue that is preventing the user from checking out. I now can resolve multiple issues and make sure my customers are able to browse and buy, saving my business. Here you have all the resources you need to debug hard to reproduce issues and the end user's device. In the Network tab you'll see the requests ordered chronologically. The N+1 query is this 13 second long request and down below we can see the 500 response that happen at checkout because we've instrumented Sentry across our entire stack we are now able to leverage spans to trace issues from a poor user experience in our mobile app down to slow database queries. Sentry makes debugging much easier with our detailed and contextualized traces. The replay video in combination with the rich debugging context I shared removes all the guesswork from being able to repro your issues, understand the user impact and ultimately being able to solve them. Thank you all for watching, and be sure to check out our Launch Week blog posts for more details. Bye for now.