Collect memory peak usage for CI jobs by default
Everyone can contribute. Help move this issue forward while earning points, leveling up and collecting rewards.
Proposal
Inside a Docker container, one can use /sys/fs/cgroup/memory/memory.max_usage_in_bytes to access the information about peak job usage. You can use metric reports to collect this information for every job.
But I think it would be useful if GitLab would be doing that by default and show it for all jobs. When one has multiple runners with multiple configurations, it is useful to know which jobs to run on which runners and with which memory limits. This can help optimize usage.
By collecting this by default, once you start optimizing, you already have historic data for your jobs. Moreover, having that information available could enable GitLab.com to determine which tiers of common runners to provide to users, too.