Issue with Netgate - routing issue and device not rebooting
-
Hello, today my collegue contacted me, that some dices are not sending data to cloud. I login into pfsense (26.03) and realized that is slower then normaly. But no high CPU usage on dashboard. I tried to make packet capture, but there was nothing from devices he mentioned. On this device we have main fibre PPOE and I setuped failover with 4G of different provider. So I was thinking it is related (and I need help here, to setup failover properly).
But then I decided to test if reboot will help, so I tried reboot device.
Unfortunately it didnt. In system logs I get this:
Jun 2 08:42:39 nginx 2026/06/02 08:42:39 [crit] 21446#101191: *103 SSL_write() failed (SSL: error:8000000D:system library::Permission denied:tls_retry_write_records failure) (13: Permission denied) while processing HTTP/2 connection, client: 172.27.30.2, server: 0.0.0.0:10433 Jun 2 08:42:39 nginx 2026/06/02 08:42:39 [error] 21446#101191: *103 upstream timed out (60: Operation timed out) while reading response header from upstream, client: 172.b.b.2, server: , request: "POST /diag_reboot.php HTTP/2.0", upstream: "fastcgi://unix:/var/run/php-fpm.socket", host: "172.b.b.1:10433", referrer: "https://172.b.b.1:10433/diag_reboot.php"and then:
Jun 2 08:48:17 kernel sonewconn: pcb 0xfffff801106f6a80 (x.x.x.x:53 (proto 6)): Listen queue overflow: 193 already in queue awaiting acceptance (44 occurrences), euid 0, rgid 0, jail 0 Jun 2 08:47:13 kernel sonewconn: pcb 0xfffff801106f6a80 (x.x.x.x:53 (proto 6)): Listen queue overflow: 193 already in queue awaiting acceptance (104 occurrences), euid 0, rgid 0, jail 0 Jun 2 08:46:13 kernel sonewconn: pcb 0xfffff801106f6a80 (x.x.x.x:53 (proto 6)): Listen queue overflow: 193 already in queue awaiting acceptance (1 occurrences), euid 0, rgid 0, jail 0What do you think yout it? Physical reboot helped. I was then even not able to connect vy VPN, most of services stopped to work.
-
Any error logged after the reboot?
Are you booting from eMMC?
-
@stephenw10 I was ot on site, is there way how to get it? from which file? There is installed SSD and it is running from it. It is model 4200.
-
You should be able to get all the recorded logs from the device if you can access the GUI. Unless it booted into an old BE?
-
@stephenw10 Can I somehow securely upload it here, so it will be not public?
-
You can upload logs here: https://nc.netgate.com/nextcloud/s/iDXZSgWKP86eqLD
-
Sorry, I needed tor eboot it more times. Now I see this error in log:
sonewconn: pcb 0xfffff8011b2a2000 (172.27.25.1:53 (proto 6)): Listen queue overflow: 193 already in queue awaiting acceptance (27 occurrences), euid 0, rgid 0, jail 0What it means and why it is happening? Am I able to resolve it without reboot of whole console? And it can be caused because we setup WAN failover ?
-
@stephenw10 I just uploaded files. We had to reboot again. Can you please check? Can be the problem also related to fact, that primary connection is PPOE?
-
What is 172.27.25.1 there?
It looks like your gateway alarms are triggering sate killing that leaves connections unable to complete until it exhausts something.
You probably need to tune the gateway monitoring better for the new WAN and possibly set the failover state killing behaviour differently.
