Instana now comes with a new landing page, which helps the user focus on key entities in various areas. The new landing page comes with four five tiles which can be reordered depending on which areas are more important.
- Applications: Shows the top 5 applications ordered by calls
- Websites: Shows the top 5 websites ordered by page views.
- Infrastructure: Based on the selected type, the infrastructure tile will show the top 5 hosts, containers or processes, ordered by CPU usage.
- Platforms: This list collects entities from various platforms like Kubernetes clusters, Cloud Foundry applications and vSphere data centers.
- Events: Shows incidents and issues that occurred in the configured time.
Our custom event system for website monitoring is now generally available after being in beta for several releases. Custom events enable reporting about non-standard activities, important interactions and custom timings to Instana. This can be especially helpful when analyzing uncaught errors (breadcrumbs), reporting information about WebSocket connections, and tracking additional performance metrics.
Complementary to our custom event system, we now automatically collect markers and measures made via the user timing API. Markers and measures are automatically translated into custom events. This means that the user timing API can be used as a vendor-neutral way to report custom events to Instana.
To determine user impact, we now use the user ID provided via the user API (preferred) or the session ID provided by opting into our session tracking. This opens up user impact insights for a much wider audience, e.g. websites that work without sign-ins and without server-side sessions.
Instana now integrates with DB2, so you can monitor performance of your DB2 instances alongside related applications, and the infrastructure that runs it all. This includes a built-in dashboard that displays key metrics about your instances’ availability, connections, query rates, and more. For more information, see the DB2 docs.
- Fixed uncaught errors in the Kubernetes service dashboard on the infrastructure map.
- Fixed back button interaction in several paginated views.
- Fixed parsing traces with unexpected parent-child relations and showing a warning in the trace detail view when this appears.
- Fixed CRI-O container naming by including the namespace and pod name when running in container-orchestration system.
- Fixed overly aggressive time skew adjustments.
- Added the ability to filter for time skews.
- Added health information to the websites overview.
- Fixed cases in which the page load visualization showed overlapping events.
- Fixed that previous baseline/threshold has been still shown for a short period when setting up an alert.
- Fixed that the UI offered options to create, delete, or edit an alert even though the user lacked the respective permission.
- Aligned the behavior of the JS errors count alerting to the related chart. Now it counts every error and not the number of erroneous page-loads.