HPCPerfStats
HPCPerfStats copied to clipboard
TACC
HPCPerfStats is an automated resource-usage monitoring and analysis package.
Since taccstats has the data required for a roofline plot, we should add it to the plots displayed.
We should explore data aggregation to get our sample times much lower. It is my thought to keep the data-points at 5 minutes, but store average/min/max/std over 10 samples (30...
Remove any statistic calculation from the javascript and keep feature parity.
We should be collecting cgroups and get the same data that is in this cgroup_exporter, we can use this data to support nodes with multiple jobs at the same time....
Currently we collect everything at a node-level. We need to examine what metrics can be split out (on a core or socket basis) and what is not able to be...
All command line tools should have a flag that can set the ini for instance ./command -i /wjerever/taccstats.ini This should be done once we have consolidated all the flags from...
We need to log the important messages and failure codes. This needs to be done before we move to systemd completely.
This PR has the new config parser library, along with updates to the ini file, and all files that read from the config parser.
Because TACC Stats is a lot of discrete scripts that run at different times, having consistency and one place to change things is crucial, especially to standardize interfaces and remove...
Add duplicate value checking, right now if there is a duplicate it stops and doesn't proccess the rest of the file. Please make the algorithm skip anything that is duplicate...