Hi - since a few days, I get errors on a quite old HA install - both, from GUI and from command line.
The error is on both nodes and I tried different mirrors, all with the same result.
configctl firmware check
OK
opnsense-update -M
https://mirror-opnsense.serverbase.ch/FreeBSD:13:amd64/23.7
pkg update -f
Updating OPNsense repository catalogue...
pkg: https://mirror-opnsense.serverbase.ch/FreeBSD:13:amd64/23.7/latest/meta.txz: Connection refused
repository OPNsense has no meta file, using default settings
pkg: https://mirror-opnsense.serverbase.ch/FreeBSD:13:amd64/23.7/latest/packagesite.pkg: Connection refused
pkg: https://mirror-opnsense.serverbase.ch/FreeBSD:13:amd64/23.7/latest/packagesite.txz: Connection refused
Unable to update repository OPNsense
Error updating repositories!
Doing a ping or curl in the same session works without problems - e.g.:
curl "https://mirror-opnsense.serverbase.ch/FreeBSD:13:amd64/23.7/latest/packagesite.pkg"
The logs show nothing special - except that the upgrade fails.
configd.py 93027 - [meta sequenceId="393"] [bb7ec313-3623-4a36-85bb-9fce6a4d4600] Script action failed with Command '/usr/local/opnsense/scripts/firmware/query.sh remote ' returned non-zero exit status 1. at Traceback (most recent call last): File "/usr/local/opnsense/service/modules/actions/script_output.py", line 44, in execute subprocess.check_call(script_command, env=self.config_environment, shell=True, File "/usr/local/lib/python3.9/subprocess.py", line 373, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '/usr/local/opnsense/scripts/firmware/query.sh remote ' returned non-zero exit status 1.
Pick a different mirror?
Quote from: Patrick M. Hausen on January 17, 2024, 10:53:07 PM
Pick a different mirror?
Thanks for your response, but ...
Quote from: chris888 on January 17, 2024, 10:35:00 PM
(...) and I tried different mirrors (...)
and all failed with the same error message.
Sorry, I concentrated on the code snippets while reading and missed that.
Is there another firewall system in front of this setup? If not, try to tcpdump on the WAN interface and find out who's sending the ICMP unreachable messages.
For me, it doesn't look like a connectivity issue. The same connection with CURL is fine.
I also tried to avoid possible issues with IPv6 by using pkg -4 in the console, but this didn't help.
Does anyone know what causes the mirrors to answer with "Connection refused"?
Connecting with curl from the firewall or any webbrowser from clients in the internal network to the mirrors works without any issues.
I finally found the problem: there was an orphaned proxy setting in /usr/local/etc/pkg.conf