You don’t know how many personalities the #AWS #CLI has until you double-check your #disasterrecovery vendor’s work during a failover test. But hey, at least now I have a pile of #shellscript that gives me strong signals.

That’s all well and good for the #infrastructure, but we may have impeded ourselves when we gave the unproven disaster recovery environment production-like (i.e. restrictive) permissions. Diagnosing why the workloads aren’t starting up properly is unnecessarily difficult.