ntpclient: stucks after a couple calls in a row, causes hotplug script to stuck, other hotplug scripts are not called anymore

[MichaelUray]

@thess

Environment: VMware OpenWrt 19.07.3, r11063-85e04e9f46 ntpclient - 2015_365-2 - NTP client for setting system time from NTP servers.

Description: Multiple calls in a row of /usr/sbin/ntpclient -c 1 -p 123 -i 2 -h 0.openwrt.pool.ntp.org cause the ntpclient to stuck. This further causes that other hotplug scripts get not executed anymore, might cause security issues since it prevents the execution of firewall hotplug scripts. I am wondering if the ntpclient package is still required, since an ntpclient comes with the current version of OpenWRT, perhaps the package should get removed.

Please look for more information there in the forum. https://forum.openwrt.org/t/hotplug-iface-script-gets-not-executed-on-boot-up/67312/48

[wulfy23]

while it is a valid issue... your underlying network settings and / or additional scripts / ntpclient settings played a role in the triggering this behavior. typical network settings, sane ntpclient settings and default codebase make this far less reproducible ( -5% vs your 50+% )

essentially... ntpclient ( iface/20-ntpclient check_server() <- NTP_RUNNING ) hangs occasionally, ( likely cause by underlying network / repeat calls ) with two failure modes... one shows no output ( -d lists 'waiting...' ) the other with the message 'rejected packet: LI==3'..., rapid succession of the command above can trigger these on a default system.

[idfxken]

This issue is still present. Was debugging a new master build for a few hours (fritzbox with VDSL loaded from PR 11728). No weird or special settings. Figuring out why my script wouldn´t work.

Moving /etc/hotplug/iface/20-ntpclient and killing the waiting processes, made everything run in a batch. It is definitly reproducible on master.