Load checks before the e05 study (rule from the task: wait up to 10 minutes, checking every 2 minutes, while the 1-minute load average exceeds the 16 logical CPUs, then run anyway and label the run).
10:41:23 UTC  hoststate.py: load 65.2 50.9 53.6 (recorded in this session's transcript, not in a file)
10:50 to 11:00 UTC  five checks two minutes apart were made by a waiting script; the script hit its own 10-minute limit before it wrote this file, so those five values were not saved. The loop stops early as soon as a check is at or below 16; it did not stop early, so at least the first four of those checks were above 16. The fifth value is unknown.
2026-11-21T11:00:40+00:00  load 46.36 66.37 63.85 (final check; the study starts now, labelled "shared, loaded host")
