JIRA Issue: KIEKER-1651 Configure queue "weight" instead of queue size
Original Reporter: Nils Christian Ehmke
Hi,
This is something we noticed during the usage of Kieker in our application:
Currently it is possible to configure the queue size for Kieker. This, however, is usually not a suitable metric in practice. Usually we are more interested in saying something like: Okay, Kieker is allowed to have 2 GB of heap space. The current workaround would be to take the used records into account and would determine some kind of average record size in order to calculate the estimated queue size. It would be better if this could be done by the queue or the writer. After all, the writer (and the writer thread) could exactly know how many bytes are currently in the queue.
JIRA Issue: KIEKER-1651 Configure queue "weight" instead of queue size Original Reporter: Nils Christian Ehmke
Hi,
This is something we noticed during the usage of Kieker in our application:
Currently it is possible to configure the queue size for Kieker. This, however, is usually not a suitable metric in practice. Usually we are more interested in saying something like: Okay, Kieker is allowed to have 2 GB of heap space. The current workaround would be to take the used records into account and would determine some kind of average record size in order to calculate the estimated queue size. It would be better if this could be done by the queue or the writer. After all, the writer (and the writer thread) could exactly know how many bytes are currently in the queue.
Thank you and best regards
Nils