You are correct and you can look at the following KB article:
In normal cases, the maximum space utilized for transactional data storage in Panorama/hedzup/MN/userdata/captures/continuous will correspond to the value specified in the configuration used for this application under 'Disk Limits/'Maximum Disk Usage, multiplied by the number of jvms or appdomains instrumented on the system. The default is 500mb per jvm or appdomain.
So this is indeed per instance being instrumented.
As Marc confirmed, the "Maximum Disk Used" setting is evaluated per process that is assigned to a given configuration. The help topic for the "Configuration Settings" was changed recently to try and emphasize that point:
Maximum disk used limits the disk space that trace files for a single monitored process can use. If trace files for any single process exceeds Maximum disk used, AppInternals deletes the oldest files in the subdirectory for the process on the agent system (<installdir>\Panorama\Hedzup\mn\userdata\captures\continuous\<hashIDforProcesses>).
Note: Total Disk Usage May Be Higher
Because Maximum disk used applies separately to the subdirectory for each monitored process, the combined disk usage by trace files for multiple processes will be higher. For processes controlled by the same configuration, the highest possible total usage will be the Maximum disk used value multiplied by the number of processes.
Note that trace files remain on the agent even after they are copied by the analysis server. Trace files accumulate and consume disk space until their subdirectory exceeds the Maximum disk used setting.
Unlike Maximum disk used, Minimum disk free applies to the entire disk where the agent software is installed. If the available disk space falls below Minimum disk free, AppInternals stops tracing for the processes. To resume tracing, you must change the setting or delete files so the available space complies with the setting.
Thanks for the information, But it would be nice to clarify that in the popup exclamation mark.
I've seen machines with over 50 instances and if you hit 'Instrument all' You will need at least 25Gb free space where the agent is installed. So Its crucial to know :-)
It would also be nice to know which files in the (<installdir>\Panorama\Hedzup\mn\userdata\captures\continuous\<hashIDforProcesses>). that is safe to delete.
one more nice to have feature is that you could change the config files on the fly without needing to restart the application that it currently monitor.
Having the "Save" operation for Configuration Settings warn about the potential disk usage is a good idea!
At least, I think the label for the "Maximum disk used" setting should be clear that it's a max for each instrumented process. I submitted a bug for both suggestions, "Configuration Settings: no warning about potential disk use by traces (customer complaint)".
I don't know the answer about whether there is a way to safely delete files in the trace subdirectories. I sent mail to the AppInternals engineering folks and will report back if I get an answer.
Not having to restart the application so often is a long-standing request. Peco Karayanev
can you comment on its status?