No worries. It sounds like the only path forward for diagnosing things would be to do a complete capture of all the network traffic happening on that node at the failure point (likely to be a substantial amount, ugh). Then go through the analysis process to work out just wtf was happening at the network level, down to the individual packet sequences at the critical time.
Super not sure it'd be worth the effort.![Frown :( :(](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
Super not sure it'd be worth the effort.