Node failure using ROS diagnostics framework
I've been reading about the ROS diagnostics framework in order to add some basic frequency checks and overall system monitoring to our system. I was hoping to get some clarification on the following question:
- What would happen if a node were to crash completely?
Based on the example here it seems like most of the calculating and analysis of publish frequency is done in the node itself with the help of diagnostic_updater. If the node were to crash, it could no longer publish this data to /diagnostics.
Is that the main purpose of the diagnostic_aggregator? to sort of "watch over" all the nodes publishing their own diagnostic data and notify of any crashes? I read through the wiki page but it wasn't entirely clear to me. Or is a crashed node considered a "stale" node? (I'm not entirely sure what "stale" means)
Any clarification would be greatly appreciated! I hope I didn't miss something obvious in the docs.
