Closed fabianfrz closed 7 years ago
I checked the log, which says the following:
Mar 9 14:27:24 opn unbound: [24343:0] error: cannot chdir to directory: (No such file or directory)
So it may be a race condition in the boot process.
Now after booting, this one shows up in the log:
opnsense: /usr/local/etc/rc.bootup: The command 'chroot -u unbound -g unbound / '/usr/local/sbin/unbound-anchor' -a '/var/unbound/root.key'' returned exit code '1', the output was ''
Can you give more context than the error itself? There is some kind of ordering issue, but I have no idea where and cannot reproduce.
It happens when the device is booting. The service can be started later without issues. I think this is because the service is started before the configuration is generated
Yes, but with some more log messages we can find the call spot. ;)
sure:
Mar 11 17:27:42 opnsense: /index.php: Successful login for user 'root' from: 10.230.231.9
Mar 11 17:26:46 kernel:
Mar 11 17:26:46 kernel:
Mar 11 17:26:45 kernel:
Mar 11 17:26:33 squid[36296]: Squid Parent: (squid-1) process 36475 started
Mar 11 17:26:33 squid[36296]: Squid Parent: will start 1 kids
Mar 11 17:26:23 squid: No valid signing SSL certificate configured for HTTPS_port 127.0.0.1:3129
Mar 11 17:26:13 ftp-proxy[23677]: listening on 127.0.0.1 port 8021
Mar 11 17:26:08 configd.py: generate template container OPNsense/Syslog
Mar 11 17:26:08 configd.py: generate template container OPNsense/Sample/sub2
Mar 11 17:26:08 configd.py: generate template container OPNsense/Sample/sub1
Mar 11 17:26:07 configd.py: generate template container OPNsense/Sample
Mar 11 17:26:06 configd.py: generate template container OPNsense/Quagga
Mar 11 17:26:03 configd.py: generate template container OPNsense/Proxy
Mar 11 17:26:01 configd.py: generate template container OPNsense/Netflow
Mar 11 17:26:01 configd.py: generate template container OPNsense/Macros
Mar 11 17:26:00 configd.py: generate template container OPNsense/IPFW
Mar 11 17:25:58 configd.py: generate template container OPNsense/IDS
Mar 11 17:25:57 configd.py: generate template container OPNsense/FtpProxy
Mar 11 17:25:57 configd.py: generate template container OPNsense/Cron
Mar 11 17:25:56 configd.py: generate template container OPNsense/Captiveportal
Mar 11 17:25:55 configd.py: generate template container OPNsense/Auth
Mar 11 17:25:54 sshlockout[12808]: sshlockout/webConfigurator v3.0 starting up
Mar 11 17:25:54 configd.py: [5572e883-33aa-459b-850a-6723cdeb055a] generate template *
Mar 11 17:25:54 kernel: done.
Mar 11 17:25:52 root: /etc/rc.d/hostid: WARNING: hostid: unable to figure out a UUID from DMI data, generating a new one
Mar 11 17:25:51 configd.py: generate template container OPNsense/Syslog
Mar 11 17:25:50 kernel: done.
Mar 11 17:25:50 configd.py: [2659fcf9-b91b-4810-bed1-6a5b52d205f1] generate template OPNsense/Syslog
Mar 11 17:25:47 kernel: ...done.
Mar 11 17:25:47 kernel: .
Mar 11 17:25:46 kernel: .
Mar 11 17:25:44 sshd[46582]: Server listening on 0.0.0.0 port 22.
Mar 11 17:25:44 sshd[46582]: Server listening on :: port 22.
Mar 11 17:25:44 kernel: done.
Mar 11 17:25:43 kernel: done.
Mar 11 17:25:42 kernel: done.
Mar 11 17:25:36 opnsense: /usr/local/etc/rc.bootup: The command 'chroot -u unbound -g unbound / '/usr/local/sbin/unbound-anchor' -a '/var/unbound/root.key'' returned exit code '1', the output was ''
Mar 11 17:25:36 kernel: done.
Mar 11 17:25:36 kernel: done.
Mar 11 17:25:36 opnsense: /usr/local/etc/rc.bootup: ROUTING: setting IPv4 default route to x.x.x.x
Mar 11 17:25:35 lighttpd[94344]: (log.c.217) server started
Mar 11 17:25:35 kernel: done.
Mar 11 17:25:35 kernel: .done.
Mar 11 17:25:35 kernel: ..
Mar 11 17:25:35 kernel: .
Mar 11 17:25:33 kernel: pflog0: promiscuous mode enabled
Mar 11 17:25:32 kernel: done.
Mar 11 17:25:32 opnsense: /usr/local/etc/rc.bootup: Resyncing OpenVPN instances.
Mar 11 17:25:32 kernel: ath0_wlan1: ieee80211_new_state_locked: pending INIT -> SCAN transition lost
I am guessing the /usr/local/sbin/unbound-anchor invoke changed return values, at least it doesn't look like it's not working. Some cleanups are incoming, but they shouldn't "fix" the "problem". Step by step...
Works for both /var MFS on and off after reboot from here. What else could we try?
Mine is an APU 1d4 with an SD card for the OS. It may happen because of slow memory
So /var MFS? If the file system / mkdir returns from creating /var/unbound it should make sure it's available...
This is the setting on this device:
Mounts:
# mount
/dev/da0s1a on / (ufs, local, soft-updates)
devfs on /dev (devfs, local, multilabel)
tmpfs on /tmp (tmpfs, local)
tmpfs on /var (tmpfs, local)
devfs on /var/dhcpd/dev (devfs, local, multilabel)
Devices
camcontrol devlist
<Multiple Card Reader 1.00> at scbus6 target 0 lun 0 (da0,pass0)
I hope any of this configuration elements can help. For now I will disable this tmpfs so the race condition will not prevent it from booting.
@fabianfrz Can we close this? It looks like a local issue that is impossible to address...
@fichtner yes, I think I can reopen it, if it appears again
After booting the DNS-Resolver seems to be not starting on
17.1.2
It runs after