Hi,
We have a writer that sends a message to 5 readers at 1hz. Every few hours we see on_sample_lost being called on all of the readers. It's only called once, but each time it says that 7 samples were lost. Nothing unusual is reported on the writer. I'd like to investigate the cause and would appreciate any suggestions.
- We use a deadline of 2 seconds for both readers and writers
- KEEP_LAST_HISTORY_QOS is 16
- Reliability is set to RELIABLE_RELIABILITY_QOS for both readers and writers
- The publish mode is ASYNCHRONOUS_PUBLISH_MODE_QOS
- We're using UDP multicast
Any suggestions would be appreciated
Thanks,
Fidel
In the end this was caused by a low level HW interrupt that was stalling the entire OS for up to 4 seconds at a time