Issue All servers dying upon reboot: Cannot assign requested address

[ShoneA]

I've been experiencing numerous issues recently with all our plesk servers. Four servers, one with a different host server, OS and Plesk version entirely.

---------------

What happens:

- Reboot server, upon returning nginx fails to start because it can't bind to an ip "Cannot assign requested address".

- On most servers this is the ipv6 address. Even if I have some sort of networking issue, Plesk doesn't let me fix the issue very easily. My fix is: Remove IP from every customer manually (ugh), then reboot. Re-reading IP's doesn't fix anything. Removing the ipv6 address doesn't work because it's the "last card on the interface" or whatever that message is.

- On one server, it just garbled all the IP's. It was failing on an ipv4 address and I had to re-read IP's to fix it.

---------------

This issue started with the ipv4 on one server, and since then every other server I reboot for the first time has issues with the ipv6 IP.

IPv6 is working:

ping6 ipv6.icanhazip.com
PING ipv6.icanhazip.com (2604:7780:200:305:f816:3eff:febd:30b5): 56 data bytes
64 bytes from 2604:7780:200:305:f816:3eff:febd:30b5: icmp_seq=0 ttl=52 time=67.499 ms
64 bytes from 2604:7780:200:305:f816:3eff:febd:30b5: icmp_seq=1 ttl=52 time=66.905 ms

I don't know what happened. Maybe all my servers updated to a version with a bug? Maybe the problem is ours? Even still, the solution Plesk gives me is difficult. Is there any easier ways or a known fix or anything anyone knows that can help me out?

Server 1: OpenSuSE 13.1 Plesk 12.0.18 Update #92 (not sure if it had problems)
Server 2: OpenSuSE 13.1 Plesk 12.0.18 Update #92 (problems the day before yesterday)
Server 3: OpenSuSE 13.1 Plesk 12.0.18 Update #92 (problems yesterday)
Server 4: OpenSuSE Leap 42.1 Plesk 12.5.30 Update #44 (problems this morning)

 

[UFHH01]

Hi ShoneA,

pls. consider to use the SEARCH option at the forum, or as well at the Plesk knowledge - base , because your issue has been discussed quite often in the past.

=> https://talk.plesk.com/search/search?&keywords="Cannot+assign+requested+address"+"bind"

=> https://kb.plesk.com/?q="Cannot ass...g=de&qLangEn=0&displayNumber=20&displayPublic


If you still experience issues, or can't investigate the root cause on your own, pls. consider to post some depending log - file - entries ( i.e.: boot.log / syslog / messages / dmesg - or what ever the relevant log - file on your operating system is... ). In addition it could help to post depending configuration files from your network - configuration(s). Consider to post the used suggestions/resolutions from threads and/or KB - articles, so that other users willing to help you know, what you already did to solve your issue.

 

[trialotto]

I've been experiencing numerous issues recently with all our plesk servers. Four servers, one with a different host server, OS and Plesk version entirely. ... (abbreviated) .......

@ShoneA

It indeed seems the case that your reboot issues (Nginx fails to start) are related to IPv6 settings.

The current Nginx version is compiled with the " --with-ipv6 " flag, so there should be no issues with the Nginx binary, but you should check that (run the command: nginx -V).

The default Nginx config files, shipped with Plesk, can be missing IPv6 related directives, in order to allow for simultaneous IPv4 or IPv6 support.

The general trick is to add the line

listen [::]:80;

in the server block(s) that handle http requests and to add the line

listen [::]:443 ssl;

in the server block(s) that handle https requests.

The [::] part is an implicit IPv6 reference, you should consider to make it an explicit reference with the exact IPv6 address(es).

Note that you should place the above mentioned lines below the normal listen directives.

Now, this can be quite cumbersome, since the custom configuration of the added directives are in danger of being overwritten.

In order to tackle that "persistence" issue, one can consider to create a custom Nginx template (used for rendering Nginx config files) OR change the IP setup of your server.

In general, I would recommend to easy method: disable IPv6 if it is not required.

However, in your case, this would probably not be a solution, since the root cause of the problem is very likely to be more complicated.

Please try to add the IPv6 related Nginx directives and see whether the issues still occur when rebooting the machine.

Regards.........