PR #178 intentionally makes admin audit writes best-effort so admin actions stay available even if the audit insert fails.
Follow-up:
- add an explicit metric/counter for admin audit insert failures
- make it easy for ops to alert on systematic audit-log degradation without relying on log scraping alone
- keep the availability-over-durability behavior, but improve visibility when it is degraded
This is non-blocking observability follow-up work.
PR #178 intentionally makes admin audit writes best-effort so admin actions stay available even if the audit insert fails.
Follow-up:
This is non-blocking observability follow-up work.