WebGUI inaccessible locally, through TS and multiple browsers.
-
Re: WebGUI crashes with 50x when opening dashboard
8200 with 6 months in production. WebGUI inaccessible yesterday. Left it overnight hoping it would resolve. Log filled with "check_reload_status 653 Could not connect to /var/run/php-fpm.socket"
Restarted firewall to the annoyance of many. WebGUI now works fine.
Downstream Ubiquiti routers show increases in ping delay from when WebGUI wasn't working.
8200 running 25.07.1-RELEASE has been stable until yesterday. Any thoughts?
-
@almostmagic did ssh work? console? You can restart the webgui from there without having to bounce the whole thing.
To be honest I don't think I have ever had to bounce the whole box other than when upgrading it. Not like I haven't seen services crash, like unbound or the gui or vpn, dhcp etc. over the years - but I don't recall ever having to bounce the whole thing to recover a service.
-
@almostmagic said in WebGUI inaccessible locally, through TS and multiple browsers.:
Downstream Ubiquiti routers
-
@johnpoz said in WebGUI inaccessible locally, through TS and multiple browsers.:
@almostmagic did ssh work? console? You can restart the webgui from there without having to bounce the whole thing.
To be honest I don't think I have ever had to bounce the whole box other than when upgrading it. Not like I haven't seen services crash, like unbound or the gui or vpn, dhcp etc. over the years - but I don't recall ever having to bounce the whole thing to recover a service.
Thank you for your quick response. I didn't have SSH enabled. I will use console on the next problem. I did narrow the issue to a failing SFP adapter that was flapping . . causing dpinger to get overly excited, which must have caused the WebGUI to fail also. This is another issue I've noticed . . when one gateway fails, it causes some adjacent gateways to fail even when they are healthy. I have gateways configured with different failure thresholds- FO connections (I have 3) are set to the most rigorous 1 sec ping, default the rest. Failover Starlink (4 of those) is set to a more leisurely 2 sec ping, higher packet loss/delay because if FO fails AND Starlink, well there isn't any internet so why ping so frequently? VPN's also have ping standards set to every 2-5 secs as they are also not as high a priority. Do varying ping standards mess with dpinger? I was assuming that setting secondary gateways to ping every 2-5 sec would reduce the load on dpinger and make it more stable.
-
@tinfoilmatt said in WebGUI inaccessible locally, through TS and multiple browsers.:
@almostmagic said in WebGUI inaccessible locally, through TS and multiple browsers.:
Downstream Ubiquiti routers
Ha! Well a year ago I wouldn't have even dreamed I'd be saying that, but for home users downstream from me it makes it a lot easier for them to manage their own networks. I love Netgate, but very few of my clients would want to bother learning pf. Did I mention the DPI?
