Provide graphical monitoring that shows the state of all of the buffers and queues in the system
When data taking stops, or is running more slowly than expected, it is not always clear what the cause of the problem is. It could be that the writing of the data to disk has significantly slowed down and that is exerting backpressure through the system. Or, it could simply be that the trigger rate (and therefore the number of fragments and events that are input to the system) has dropped.
To help understand what is going on in situations like this, it would be helpful to have a graphical monitor that shows the states of all of the buffers and queues in the system. Such as a display should provide sufficient detail so that a DAQ expert could use it to diagnose problems. And, it should also have sufficient clarity that a shifter could use it to get a sense of the general area of a problem, when a problem occurs.
This Issue asks for a particular solution to the problem of understand where the bottlenecks in the system are at a given time, but there may be other solutions that would work just as well. If so, those should be proposed also.
A companion display that shows the state of each of the processes in the DAQ system may also be helpful.
#1 Updated by Eric Flumerfelt over 3 years ago
- Project changed from artdaq to artdaq Utilities
- Category set to 431
- Target version deleted (
- % Done changed from 0 to 10
I have a very basic version of this working in artdaq-node-server. With TransferPlugin-based data flow, I think I might have enough directed stats to make a neat GUI.