What Happens During an Application Crash? Does the Session Ever Get Logged as Stopping?
What Happens During an Application Crash? Does the Session Ever Get Logged as Stopping?
If an application crashes, it never gets a chance to call the SDK to stop, therefore initially the session is not logged, however Usage Intelligence has a built in mechanism that can detect when a session has not reported back data to the server within a week, and at that point the session is marked as stopped for reporting purposes.
Transcript
Well what happens if there's an application crash and so as a result I never get to call RUI stop does that session ever get logged? Did it ever stop? What what happens to the data and sort of what happens in that situation?
We've actually designed the technology to compensate for this and actually we have an interesting mechanism in place to be able to track this. So basically if a crash occurs what's going to happen is the call to stop SDK will never get called therefore the current session actually doesn't terminate because that stop SDK call will never be made but the server has a safeguard in place to still collect this data and basically what happens is if no data from this session is collected for one week so we never see that session terminate for a period of one week the server will mark this session as a session timeout and then after one week when the user relaunches the product that remaining data from that time out session will be collected.
We can actually show you what happens in the dashboard for that case if a session time out occurs. If we take a look at our user flow report again we tag that as an event when a session timeout occurs you can select this session timeout as a hotspot within the user flow report and at that point you'll be able to visualize the sequence of events that lead to this session timeout so this could actually be like a good benefit here because you'd be able to see under what instances am i seeing a session timeout get triggered. I can see here maybe this flow in particular is leading to or indicative of a crash within my product or is leading to some kind of unexpected behavior that's causing the session timeout to occur. So even in the case of an application crash there's still some data that we're going to collect in at later point that you'll actually be able to really visualize the dashboard to help to diagnose what might be causing that crash.
