Resolved ipv6 stops working after 10-15min

[KNARZ]

I want to fine tune some server and from my observation I run in a ipv6 Problem.
If reboot the server everything works fine as it should but after 10-15min ipv6 stops working.

I'm not sure if it has to do with the hosting/server, configuration or the webserver (apache, nginx) at all but why does it work in the beginning and than stops?

• OS: ‪Debian 8.7‬ - amd64
• Product: Plesk Onyx 17.0.17 Update #13

I tried to rebuild the config and restartet nginx and apache but nothing changed.

I'm testing this with http://ipv6-test.com/validate.php
Results:

 

[IgorG]

Have you tried to find something useful in logs? Any results of troubleshooting?

 

[KNARZ]

Yeah I'm pretty basic - there are so many logs, that I don't know which one I should look into.

 

[IgorG]

Yeah I'm pretty basic - there are so many logs, that I don't know which one I should look into.

This article will help you - https://support.plesk.com/hc/en-us/articles/213403509

 

[KNARZ]

thanks for the info. I found some things in syslog but so fare nothing related (in this 15min timeframe) but I fixed some other minor issues.
I contacted the hoster. They said I should try it within recovery console and than check if ipv6 is working.

 

[KNARZ]

I guessed it's about psa-firewall rules. But now after a while (like 30min) all ping request timed out. - I will investigate further. - Nothing in particular in the syslog

 

[KNARZ]

Okay.
So there were two things.

1) Testing with ping was blocked by PSA-Firewall
2) Testing with the Test-Website failed because of DHCP v4 auto config after 30 minutes.

/etc/network/interfaces looks like:

# The primary network interface
auto eth0
iface eth0 inet dhcp

post-up while ip -6 addr show tentative | grep . > /dev/null ; do sleep 1 ; done

iface eth0 inet6 static
       address X
       netmask Y
       gateway Z

post-up while ip -6 addr show tentative | grep . > /dev/null ; do sleep 1 ; done

After exactly 30min IPv6 will fail. That is caused by not configured:

net.ipv6.conf.eth0.accept_ra=0

So I placed a /etc/sysctl.d/ipv6.conf
with the above line of code and restarted.
Now IPv6 works fine.

Sources that helped me by this issue:
https://www.gibts-doch-garnicht.de/ipv6-bei-manitu-und-netcup/
https://debianforum.de/forum/viewtopic.php?f=30&t=155189

 

[robetus]

Thank you KNARZ, this worked for me as well using a CentOS 7 LXC container for plesk. Much appreciated, hopefully others will find this.