To improve data collection and performance we worked on our log system. We have:
- Improved logging dead band options
- Relative change or full scale.
- Change as floating point value.
This results in decrease amount
of stored log data.
- Increased number of log files per hour
- For large systems (2000 objects and above)
logs are distributed in 16 files per hour
instead of 1.
This results in improved read performance
without significant write impact.