search for: process_plock

...ontrold.pcmk stack) where we see fnctl(F_SETLK) return ENOSYS. This happens after a node has failed/been fenced and recovered. It affects only individual mounts, not the whole system. The only code path that I can see so far that could lead to ENOSYS being returned here is from within dlm_controld:process_plocks() if disable_plocks == 1. But I can't find the tell tale log message that I think we should see (from store_plocks) if this value was set. We'll continue investigating, but perhaps this rings a bell for someone and we can shorten the latency ;-) Thanks for any insights. Regards, Lar...