We have a script that runs at 6:00am every morning to download the .nar file from our ns480, converts it to a csv and loads it into a database. The last couple weeks, the data from 6:00am to 6:00pm is missing. So if I inquire on a given day, I have the 12:00am - 6:00am, then the 6:00pm to 11:59pm.
Has anyone else experienced this? Is it a bug? Should I reboot the management server and restart logging?
The data points got changed, so the data was being rolled off the file. Changed the data logging points (real time interval and archive interval) to 60/600 and now it is working properly.
Nice job on finding the culprit. Makes sense that somewhere along the way the Archive Interval was changed. At the risk of repeating what you already know, if you don't have Periodic Archiving enabled, then you'll only be able to retrieve 156 samples from the rolling buffer. Or worded differently, an archive will only have (up to) 156 samples.
Therefore, for a single archive to have at least 24 hours, you will need to set the Archive Interval to:
24 hr * 60 min/hr * 60 sec/min / 156 samples = ~554 seconds
We have a job that runs at 07:00 everying morning that cuts a new archive, downloads the old one, converts to .csv, and loads into a database, so we can graph utilization, bandwidth, and throughput for each day.
Was working great, until the intervals got changed, but now back to working.
PLease mark your question as Answered, that makes it easier for others to find your solution to a similar problem that has been resolved.