On 5/22/2019 6:46 AM, Roderick wrote: > Any recommendations in general? Current or stable?
I've had bad luck with softupdates and OpenBSD on ESXi when the ESXi datastore is on nfs. (Encountered on ESX 5.0, 5.1, and 5.5; I must not learn from my mistakes.) From what I can tell, if the nfs datastore takes too long to respond OpenBSD thinks the disk has gone away and panics. It's not OpenBSD's fault; if a real disk stopped responding a panic is probably the best result. Probably some tuning could be done to make the ESXi nfs timeout match the OpenBSD timeout, but it's easier to just not use softupdates and match the backing disk to my performance requirements. Note that softupdates are not enabled by default.