Re: openvpn and no buffer space available (13.2-stable)

Am 21. September 2023 20:16:34 MESZ schrieb void <void@f-m.fm>:
>Hello @net,
>
>tl;dr : is there anything specific to freebsd that needs to be set
>in order for openvpn to perform well? What buffer space is ping
>complaining about?
>
>context is recent 13.2 stable, on amd64, and it's a bhyve guest.
>The openvpn client uses UDP, on tun0.
>
>The problem is that when the connection becomes heavily
>used, the client end finds that sites that would normally
>immediately load often wont; one has to sit there clicking
>multiple times to get any site to load.
>
>ping shows this:
>
>64 bytes from 93.184.216.34: icmp_seq=37 ttl=53 time=147.407 ms
>ping: sendto: No buffer space available
>64 bytes from 93.184.216.34: icmp_seq=40 ttl=53 time=174.738 ms
>64 bytes from 93.184.216.34: icmp_seq=41 ttl=53 time=119.048 ms
>64 bytes from 93.184.216.34: icmp_seq=42 ttl=53 time=169.223 ms
>ping: sendto: No buffer space available
>64 bytes from 93.184.216.34: icmp_seq=44 ttl=53 time=183.493 ms
>64 bytes from 93.184.216.34: icmp_seq=45 ttl=53 time=162.594 ms
>^C
>--- example.org ping statistics ---
>46 packets transmitted, 36 packets received, 21.7% packet loss
>round-trip min/avg/max/stddev = 118.258/156.360/184.931/19.508 ms
>
>I've ran mtu-test on the client, which gives these results
>
>Empirical MTU test completed [Tried,Actual] local->remote=[1455,1455] remote->local=[1427,1427]
>
>in /etc/sysctl.conf, I've changed these values:
>
>net.inet.udp.recvspace=524288
>kern.ipc.shm_use_phys=1
>net.inet.tcp.sendspace=524288
>net.inet.tcp.recvspace=524288
>net.inet.tcp.rfc1323=1
>net.inet.tcp.always_keepalive=0
>kern.ipc.soacceptqueue=524288
>
>in /boot/loader.conf, these:
>
>kern.maxusers="4096"
>vm.vnode_pbufs="10240"
>kern.ipc.nmbclusters="5000000"
>
>tia,

I run OpenVPN without any such problems. My sysctl.conf as well as my loader.conf are pretty much default.

I'd try what happens without vm.vnode_pbufs="10240"

Regards,
Mathias