We got carried away by building features and improving the application experience, that we forgot to announce the changes. Here are recent changes at a glance:
- New lambda view
- New invocation view
- Metric based alerting
- Error reporting
New lambda view
The new view has separate charts for invocations, errors, duration, memory usage and cost. On the metric panel, you can see the sparklines, which indicate what is happening over the selected period and clicking on them opens up the chart in full size. Picking the period is more straightforward, there are only four options at the moment: 1 hour, 1 day, 7 days and 30 days. The bottom part includes recent invocations, which can be filtered down to errors, retries, cold starts or anomalies, and all the errors that have happened with this lambda.
New invocation view
Instead of having a separate full page for invocations, it slides open from the right. You can still scroll around lambda’s invocations list and open up a new one quickly.
Now you can set up metric based alerting for any lambda or project. Just open up policies view under alerting tab and start adding the conditions. Whenever a condition fails, an incident opens and stays that way until the term passes again. For every policy, you can attach multiple notification channels, which are shared within the organization account, so that you only need to define them once.
Errors view got design updates and moved to policy-based notifications. You can choose which types of errors you want to receive for selected targets. We generated the default policies for you to receive all errors of all lambdas, but feel free to change it. At the moment we only send out notifications when a new error occurs, or it reopens. We recommend to resolve the error after bugfix or mute it when it is not important.
- Added Dashboard menu item for easier navigation. Clicking on the logo still works as well
- Search and header have a new look and feel
- API Gateways got design overhaul
- Logs now supports JSON formatting