By default, tracepusher automatically generates it's own trace and span IDs. This is perfect for the "default" usecase when someone just wants to push a single span trace to a backend for demo purposes.
However, consider this (slightly more complex) bash script:
v0.3.0 (and prior) for tracepusher would only give a single span of 2 seconds. No details of the "inner working".
The enhancement described above would allow tracepusher to generate which describes the "internals", like this:
In this case, the user (as part of the shell script) would be responsible for generating the traceid (which would become the parent trace id). Then a new span id for each subspan. tracepusher would be called once at the "top" for the parent and then n times (in this case 2) inside the loop. Inside the loop, the parent trace ID would be the same as the original trace ID.
Perhaps this requires another flag --has-parent=True?
By default, tracepusher automatically generates it's own trace and span IDs. This is perfect for the "default" usecase when someone just wants to push a single span trace to a backend for demo purposes.
However, consider this (slightly more complex) bash script:
v0.3.0
(and prior) for tracepusher would only give a single span of 2 seconds. No details of the "inner working".The enhancement described above would allow tracepusher to generate which describes the "internals", like this:
In this case, the user (as part of the shell script) would be responsible for generating the traceid (which would become the parent trace id). Then a new span id for each subspan. tracepusher would be called once at the "top" for the parent and then
n
times (in this case2
) inside the loop. Inside the loop, the parent trace ID would be the same as the original trace ID.Perhaps this requires another flag
--has-parent=True
?