Provisioning service starts before DNS, provisioning fails at boot

The provision service seems to boot up before DNS does, so it is failing to resolve the configuration file location. It will fail, then try again in 1 hour and succeed. Can the provision service be modified to only start up after DNS is working? Our workaround is to simply use an IP in the config file location, rather than our provisioning server, but thought I’d run it by the forums.

It is also odd that it says no servers found in /etc/resolv.conf initially. I wonder if it has anything to do with us setting DNS servers manually?

Here’s the syslog file:
<Tue Jan 12 17:14:47 2021> system: system reboot by
<Mon Oct 28 19:32:23 2019> LinkStatus: WAN Link Up
<Mon Oct 28 19:32:23 2019> LinkStatus: LAN1 Link Down
<Mon Oct 28 19:32:23 2019> LinkStatus: LAN2 Link Down
<Mon Oct 28 19:32:24 2019> LinkStatus: LAN3 Link Down
<Mon Oct 28 19:32:24 2019> LinkStatus: LAN4 Link Up
<Mon Oct 28 19:32:26 2019> udhcpc[4985]: udhcpc (v1.12.1) started
<Mon Oct 28 19:32:27 2019> udhcpd[5228]: udhcpd (v1.12.1) started
<Mon Oct 28 19:32:28 2019> dnsmasq[5766]: started, version 2.40 cachesize 150
<Mon Oct 28 19:32:28 2019> dnsmasq[5766]: compile time options: IPv6 GNU-getopt no-RTC no-MMU no-ISC…
<Mon Oct 28 19:32:28 2019> dnsmasq[5766]: running as root
<Mon Oct 28 19:32:28 2019> dnsmasq[5766]: no servers found in /etc/resolv.conf, will retry
<Mon Oct 28 19:32:28 2019> dnsmasq[5766]: read /etc/hosts - 2 addresses
<Mon Oct 28 19:32:31 2019> udhcpc[4985]: Sending select for 64.79.XX.XX…
<Mon Oct 28 19:32:31 2019> udhcpc[4985]: Lease of 64.79.XX.XX obtained, lease time 28800
<Mon Oct 28 19:32:36 2019> provision[11217]: provision v1.2 start
<Mon Oct 28 19:32:36 2019> provision[11217]: start to check config file
<Mon Oct 28 19:32:37 2019> udhcpd[11551]: udhcpd (v1.12.1) started
*<Tue Jan 12 17:15:44 2021> provision[11217]: Update configuration fialed, retry 3600 s later
<Tue Jan 12 17:15:44 2021> provision[11217]: start to check command file
<Tue Jan 12 17:15:45 2021> dnsmasq[5766]: reading /etc/resolv.conf
<Tue Jan 12 17:15:45 2021> dnsmasq[5766]: using nameserver 8.8.8.8#53
<Tue Jan 12 17:15:45 2021> dnsmasq[5766]: using nameserver 64.79.XX.XX#53
*<Tue Jan 12 17:15:47 2021> snmpd[16520]: start

Hi ,

We will try to get this fixed in later releases .
For workaround you can try setting ,“Upgrade Error Retry Delay(sec)” , to a smaller value say , 300 seconds . Which is by default is 3600 seconds . And it will retry provision sooner .

2 Likes

This seems to be addressed in 4.7.3. Thanks :+1:

1 Like