commit | 58e5f7d020e7c93a645212ebeabf515b5cf81d81 | [log] [tgz] |
---|---|---|
author | Daniel Gruno <humbedooh@apache.org> | Tue Apr 30 17:27:49 2024 -0500 |
committer | Daniel Gruno <humbedooh@apache.org> | Tue Apr 30 17:27:49 2024 -0500 |
tree | 4dd3feb96b1e132ee51fd03db6bec414eaa37e9c | |
parent | 62ddf0cccaef604ce5d74f211f90bdf295ee5da5 [diff] |
Keep MAX_BUILD_SPAN hours in memory, but cut away 'jobs' after DEFAULT_BUILD_SPAN We will want to get the overall stats for the past 30 days, but not necessarily the jobs data. By removing the memory footprint for the jobs arrays, we can store the full 30 days in memory with minimal additional memory footprint, as the jobs and individual job steps can account for up to 85% of the total builds entry.
This serves as the basis for the ASF Infrastructure Reporting Dashboard, used to display various statistics and live charts on data pertaining to a wide variety of infrastructure aspects at the ASF as a unified dashboard.
WIP.