bind: cannot assign requested address

### Prerequisites

- [X] I have checked the [Wiki](https://github.com/AdguardTeam/AdGuardHome/wiki) and [Discussions](https://github.com/AdguardTeam/AdGuardHome/discussions) and found no answer

- [X] I have searched other issues and found no duplicates

- [x] I want to report a bug and not ask a question

### Operating system type

Linux, OpenWrt

### CPU architecture

64-bit ARM

### Installation

Other (please mention in the description)

### Setup

On a router, DHCP is handled by the router

### AdGuard Home version

v0.107.21

### Description

#### Installation
OpenWRT opkg

#### What did you do?
Reboot the router.

#### Expected result
AdGuard Home should start reliably after a system reboot.

#### Actual result
Fairly often, AdGuard Home errors out on reboot. Failed 3 out of 4 times while trying to gather a log file that demonstrates success for this ticket. Failure message is:

> [fatal] couldn't start forwarding DNS server: listening to udp socket: listen udp 127.0.0.1:53: bind: cannot assign requested address

#### Workaround
SSH into the router and restart AdGuard Home through init.d script. This method is 100% successful.
Alternatively, reboot again, which is sometimes successful.

#### Additional information
I've provided three log files for comparison.
1. reboot-success.log - verbose logs following a reboot, where AdGuard Home starts up successfully.
2. reboot-fail.log - verbose logs following a reboot, where AdGuard Home fails to start up.
3. init_d-success.log - verbose logs following a manual restart using the init.d script, where AdGuard Home starts up successfully.

[reboot-success.log](https://github.com/AdguardTeam/AdGuardHome/files/11362133/reboot-success.log)
[reboot-fail.log](https://github.com/AdguardTeam/AdGuardHome/files/11362134/reboot-fail.log)
[init_d-success.log](https://github.com/AdguardTeam/AdGuardHome/files/11362135/init_d-success.log)

#### Observations
I noticed that for the init.d restart instances, there is a 10-second delay between log entries for "Go to http" style messages, and the "Creating the UDP server socket" messages. For instance:

> 2023/04/30 18:36:30.996889 3586#1 [info] Go to http://[fe80::ea9f:80ff:fef0:6a84%wlan1]:8080
> 2023/04/30 18:36:41.302740 3586#34 [info] Creating the UDP server socket

This is interesting, but probably a red herring since there is no delay in either of the failed or success logs for the reboot scenario.

#### Things I've tried
* I tried delaying the start of AdGuard Home by setting "START=99" in the init.d script. ~~This did not help.~~ Actually, the first time I did this I made a mistake and forgot to disable and re-enable the service in order to get the new "S99adguardhome" symlinks in /etc/rc.d/. So, this actually does seems to work.
* Moot point now. ~~My router is using DNSMASQ as an upstream DNS, with port configured to 54, as described here: [https://forum.openwrt.org/t/how-to-updated-2021-installing-adguardhome-on-openwrt-manual-and-opkg-method/113904.](https://forum.openwrt.org/t/how-to-updated-2021-installing-adguardhome-on-openwrt-manual-and-opkg-method/113904) I thought maybe there could be a race condition between DNSMASQ coming up defaulted to port 53 after a reboot, but not seeing anything in logs to corroborate that.~~
* Moot point now. ~~The WAN network of my router is configured to use unencrypted DNS to work around NTP issues as described here: [https://openwrt.org/docs/guide-user/services/dns/adguard-home#bypassing_encrypted_dns_for_ntp.](https://openwrt.org/docs/guide-user/services/dns/adguard-home#bypassing_encrypted_dns_for_ntp)~~

#### Mea Culpa
There is at least some evidence that this issue has been reported, for instance here: [https://github.com/AdguardTeam/AdGuardHome/issues/3760.](https://github.com/AdguardTeam/AdGuardHome/issues/3760) Although I haven't found anything active for OpenWRT, hence the new ticket.

#### A question
So, it seems that inserting a delay in the startup process does help. Is there something else to investigate here to better understand the underlying race condition?