Denton Gentry

The instance created for us behaves differently: ``` root@tailscale:~# /usr/sbin/iptables -t nat -N ts-postrouting --wait root@tailscale:~# echo $? 0 root@tailscale:~# iptables -v iptables v1.8.2 (legacy): no command specified root@tailscale:~# ```...

The finding a few days ago is that my OpenVZ instance, on the same provider as reported in this bug, was able to function as an exit node for IPv4....

In nf_tables mode: ``` root@tailscale:~# iptables -L -v -n Chain INPUT (policy ACCEPT 14192 packets, 2042K bytes) pkts bytes target prot opt in out source destination 14181 2041K ts-input all...

NAT table in nftables mode: ``` root@tailscale:~# iptables -t nat -L -n -v Chain PREROUTING (policy ACCEPT 0 packets, 0 bytes) pkts bytes target prot opt in out source destination...

That is definitely a problem. Here are all of the modules loaded on my OpenVZ instance. ``` root@tailscale:~# lsmod Module Size Used by iptable_mangle 4242 -2 nft_chain_route_ipv4 4242 -2 nft_chain_nat_ipv4...

I have a Digital Ocean droplet which is set up as an exit node. It had been operating iptables in legacy mode. ``` root@dgentry-do-de1:~# iptables --version iptables v1.8.4 (legacy) ```...

If the exit node is using `--tun=userspace-networking`: In Tailscale releases prior to 1.32.1, userspace-networking did not handle TCP selective acknowledgement and suffered large performance loss from moderate packet loss. If...

If the exit node is using --tun=userspace-networking, you should update the exit node to 1.32.1 or later.

iOS was not able to do this for a long time due to lack of memory: https://tailscale.com/blog/go-linker/ It is probably feasible now: - The linker changes documented in https://tailscale.com/blog/go-linker/ went...

The node which sent BUG-6cd915a4a7be6c7ef4cec6ed7f41e35de7b9e6423c6fd9c36f0883c27d312149-20220919185020Z-be5bd846995295f1 is tagged. Was it tagged prior to the hard drive failure as well?