OPNsense Forum

English Forums => General Discussion => Topic started by: z0rk on October 18, 2019, 06:06:43 am

Title: Connection timeouts of services on LAN over WDS bridge
Post by: z0rk on October 18, 2019, 06:06:43 am

This is more of a general question to rule out OPNsens as fault domain.
 I've recently set up a new bridge. When I try to reach services (SSH, SMB, etc.) across the bridge the connection times out or if the connection is successful in the case of SMB for example folders with less then ten items take forever to load or the file manager freezes up. In the case of SSH I also get the initial connection timeout and once I connect the terminal after some time (a few minutes) of usage stops to accept input then it may work again or the terminal session just freezes up completely. Also, some of these hosts cannot be reached at all and pings fail. The hosts affected switches around. So sometimes A works sometimes B.
OPNsense is configure to use static DHCP and permanent ARP. IP addresses get assigned and arp correctly maps IP to MAC. There are no rules in play. There are no issues trying to connect between LAN services/devices on either end of the bridge (not across the bridge).
All devices are on the same subnet. The bridge is correctly configured with IP, gateway, DNS, etc. There are no connection issues with external services, internet, Netflix, etc. Throughput is excellent.
I've put in a ticket with the vendor (Engenius), but my experience with their customer support hasn't been stellar.
I hope that some network wizard my have feedback for me.

Thanks

Title: Re: Connection timeouts of services on LAN over WDS bridge
Post by: z0rk on October 27, 2019, 01:21:56 am

Bump
No takers?

I am happy to provide more details.
One thing I forgot to mention is that if I issue a continuous ping from the client to the SSH/SMB host and it succeeds after an initial delay (~30 seconds) then all issues are resolved, i.e. SSH terminal is responsive, SMB access is responsive. When I stop the ping behavior reverts back.
I would think this to be some routing issue. The only device routing traffic is OPNsense for the two subnets. I don't have any custom routes configured.