You are not logged in.
Pages: 1
Hello everyone,
After a few months of struggling to find a fix for this I thought it was time to make a post about it.
I have checked the wiki/forums but have not been able to find a post or doc with my same issue.
I have a NFS running on a RPI connected to my local network. The share is working fine however I have noticed it becomes incredibly slow when using it while I am connected to my VPN.
For example, I have a game I have saved on the NFS, without being connected to the VPN it takes 3 seconds to boot into the game. While connected to a VPN it takes close to 30 seconds to boot into the game.
I had this same issue in the past while using a SSHFS mount and was also never able to resolve the issue.
I assume this might be DNS related or that the connection is going through my VPN then connecting to my NFS but I never cared much for networking so I am unfamiliar with what data would be useful or how to collect it. If you need any additional information collected please let me know.
I am using OpenVPN + Network Manager for my VPN.
Fstab:
192.168.1.12:/mnt/nfsshare /home/user/Share nfs _netdev,rw 0 0Thanks an advance
Offline
Offline
Standard arch 64 install, also have tried both linux-lts and linux kernel as that was suggested at one time. RPI is running debian buster
Last edited by bloodlinesx (2019-11-26 13:40:55)
Offline
It worth to look at routing table. Perhaps when VPN is connected (can you tell also your provider?) all traffic goes through it and their DNS servers are being used. You can add manual route to your NFS server and use direct IP address (or add hostname records to /etc/hosts) to be sure your NFS route and lookup queries are served first
Offline
Provider I am using is NordVPN which does have their own DNS server. For the manual route I assume I would just add the hostname to the /etc/hosts with the ip and use the hostname in my fstab file correct? Similar to this:
fstab:
hostname:/mnt/nfsserver /home/user/share nfs rw 0 0 /etc/hosts:
192.168.1.12 hostname hostname Offline
You can just use "192.168.1.12:/mnt/nfsserver /home/user/share nfs rw 0 0 " - the question is where 192.168.1.0/24 is routed ("ip r", w/ and w/o VPN active)
Offline
Without VPN:
default via 192.168.1.1 dev wlp4s0 proto dhcp metric 600
192.168.1.0/24 dev wlp4s0 proto kernel scope link src 192.168.1.2 metric 600With VPN:
default via 10.8.2.1 dev tun0 proto static metric 50
default via 192.168.1.1 dev wlp4s0 proto dhcp metric 600
10.8.2.0/24 dev tun0 proto kernel scope link src 10.8.2.54 metric 50
104.140.14.42 via 192.168.1.1 dev wlp4s0 proto static metric 600
192.168.1.0/24 dev wlp4s0 proto kernel scope link src 192.168.1.2 metric 600 Offline
Trick with hosts is for setting alias for NFS server, not for tweaking routes.
Regarding your routing table. You have two default routes which leads to the so-called async routing and first default router (in VPN case) will reject queries to the 192.168.1.12. Seems like network subsystem waits for any type of query (accept/reject) and only then proceeds to the next table entry so you observe lag
Last edited by dviktor (2019-11-26 21:51:49)
Offline
Pages: 1