The Linux syscall test is a special test that compares the strace output of a wt program run with a known reference, to capture missing or extra system calls that data corruption issues. The test is currently scheduled to run once per week, and any failure captured requires an effort to tell which ticket in the list of tickets merged in the past week is the actual cause.

Considering the small runtime (~1 mins) and the criticality of problems it could catch, we should consider increasing the test frequency, to make it run for each commit, or even each PR.