Skip to content


How may levels deep from root will you populate data in POJOs?

This depends on how early in the code the said POJO was created/accessed compared to the trace point you are loading. This will mostly affect the fields of service/manager classes which are usually created on init. But even if it was created very early on but accessed recently around your trace point, it should come up. We are storing about 100k events in each slice. These are from all the threads which are active at the time. So for 1 thread, assuming about 20 threads executing through your code, that would be about 5k events for 1 thread, which would translate to about 1k lines of executions.

We are also discussing of implementing about some sort of pagination around these slices, that would allow you to go back/forward seamlessly.

Will I ever see the data in List, Set, Map datatypes?

If you do a list.get/set, map.get/set, or any container.get/set, those specific key/value pairs will be recorded.

Also we can potentially add size() in our records, which i think could be useful. If you stream over the data, lets say in a loop or a .map/.each, you should be able to see all the values(inside the closure)

Will I ever see the data in a POJO that is not under ai/overjet ?

You can potentially record everything, by instrumenting all packages (skip the i= parameter when running). You can also instrument multiple packages, with comma separated values like i=<package1>,<package2>

Do note that instrumenting everything will heavily affect your runtime performance.

Once I start playing a slice, how long will it play? I see that it stops after a method ends but the next immediate method in the execution order may or may not be available to play.

Yes, so this basically comes back to the first question. How much forward/back you can go depends where the trace point was matched in the slice. This would not be an issue when we implement pagination.