Ok, ktrace showed up that processes having problems to grab a socket. e.g. out of fds.
Question: How far can I rise kern.maxfiles having following conditions: 16G RAM total prefork 16 relayd procs openfiles-max=65536 in login.conf for relayd class. Out of my calc it is 16*65536 just to accommodate relayd. Am I safe to have 1+M of fds as kern.maxfiles ?? //mxb > 27 sep. 2017 kl. 21:34 skrev Maxim Bourmistrov <m...@alumni.chalmers.se>: > > My intention with this mail is to gather more qualitative help > to, hopefully, ever solve this or to have more info so it can be provided > to someone whom can solve this, if it is a bug. > > What I know for sure is that those boxes (dual-node setup) are exposed to a > large HTTP PUT/POST requests. > Large does not means ”put a 1G-sized stream”. Text it is. > > Of cause it is not a bug report. > No info, if any, states this. > > Let’s see what I can gather more with fstat. > Thanks. > > Br > Maxim > > > >> 27 sep. 2017 kl. 20:51 skrev Theo de Raadt <dera...@openbsd.org >> <mailto:dera...@openbsd.org>>: >> >> This probably means the process has run out of file descriptors. >> fstat might expose what is going on. If that's the case, roll your >> sleeves up and study the code. As usual, the circumstances that >> create this event are not being completely documented - I MEAN, what >> I'm seeing here isn't a bug report including a configuration file >> that reliably reproduces it. >
