Resolved Plesk stopped working all together

[maarten999]

After an upgrade in plesk and setting PHP-FHM to ngnix I cannot get in to the plesk admin anymore. After I try to log in, no login window appears but it keeps trying to connect, or connecting and going into a loop, because nothing shows.

In the console tried to repair, install, but nothing on plesk seems to work anymore. As if plesk is going into some loop somewhere. Please help.

 

[IgorG]

Need more details from logs.
/var/log/sw-cp-server/error_log, /usr/local/psa/admin/logs/panel.log and /var/log/sw-cp-server/sw-engine.log at least.

 

[maarten999]

Thanks for your help. I attached the logs. The issue happend last Friday, so 4th November. It seems that the panel.ini file is empty. Should it be in Plesk 12.5?

 

[bivaswap]

the error_log

2016/11/06 09:38:43 [error] 2587#0: *57 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 217.240.132.216, server: , request: "GET /login.php3?success_redirect_url=https%3A%2F%2Fh2191422.stratoserver.net%3A8443%2F HTTP/1.1", upstream: "fastcgi://unix:/var/run/sw-engine.sock", host: "h2191422.stratoserver.net:8443", referrer: "https://h2191422.stratoserver.net:8443/"

Try https://kb.plesk.com/en/129959

 

[maarten999]

Unfortunately:

dpkg: warning: ignoring request to remove librdbmspp which isn't installed

 

[UFHH01]

Hi maarten999,

pls. make sure, that sw-cp-server and sw-engine are active and consider to restart your services with the commands:

service psa restart ( OR: "/etc/init.d/psa stop && /etc/init.d/psa start" )

or/and

service sw-cp-server restart
service sw-engine restart​

Consider as well to have another look at your logs, to investigate ( possible ) errors/issues/problems, when restarting your services.

Plesk for Linux services logs and configuration files ( KB - article: 111 283 )​

 

[maarten999]

No issues visible in the plesk log files and no error messages on restarting all of the above services. Strange is that not only the admin does not work anymore, the autoinstall with web interface stops working after logging into it and trying to display the list of updates/installations available. Plesk repair seems to be getting into something right after entering the command plesk repair installation -n or -y.

 

[UFHH01]

Plesk repair seems to be getting into something right after entering the command plesk repair installation -n or -y.

Each repair proccess creates a pretty detailed log - file, which you should consider to investigate after you used the "Plesk repair utility". "Guessing" as "Plesk repair seems to be getting into something" is not really the right way to get to a root cause of an issue.



Can you confirm, that not any ( probably misconfigured ) firewall rule blocks the port(s)?
Can you confirm, that SELinux and/or apparmor isn't blocking?
Can you confirm, that ( IF you use Fail2Ban ), that you added your server IPs and "127.0.0.1/8" into your whitelist?
Can you confirm, that no ( possible misconfigured ) ModSecurity settings blocks?


Edit:
In addition, port "8880" ( alternative HTTP - port for Plesk ) is not reachable as well: => http://h2191422.stratoserver.net:8880

Pls. share the output from the commands:

service psa status
service sw-cp-server status
service sw-engine status

lsof | grep sw- | grep :8443
lsof | grep sw- | grep :8880

netstat -tulpn | grep 8443
netstat -tulpn | grep 8880​

Edit2:

You could as well try:

COMMANDS TO USE OVER THE COMMAND LINE - logged in as user "root":

rm -f /usr/local/psa/var/cache/*
service sw-cp-server restart
service sw-engine restart

 

[maarten999]

Probably the last solution, deletion of the cache, got me alot further in this issue. Yes, yes, yes, just connected to plesk again. Update server does not work appearantly.

Outputs of first edit are: (last command no output)

root@h2191422:~# service psa status

is running

root@h2191422:~# service sw-cp-server status

* sw-cp-serverd is running

root@h2191422:~# service sw-engine status

* sw-engine-fpm is running

root@h2191422:~# lsof | grep sw- | grep :8443


sw-cp-ser 11129 root 8u IPv4 195857 0t0 TCP *:8443 (LISTEN)

sw-cp-ser 11129 root 10u IPv6 195859 0t0 TCP *:8443 (LISTEN)

sw-cp-ser 11130 sw-cp-server 8u IPv4 195857 0t0 TCP *:8443 (LISTEN)

sw-cp-ser 11130 sw-cp-server 10u IPv6 195859 0t0 TCP *:8443 (LISTEN)

root@h2191422:~#

root@h2191422:~# lsof | grep sw- | grep :8880

sw-cp-ser 11129 root 9u IPv4 195858 0t0 TCP *:8880 (LISTEN)

sw-cp-ser 11129 root 11u IPv6 195860 0t0 TCP *:8880 (LISTEN)

sw-cp-ser 11130 sw-cp-server 9u IPv4 195858 0t0 TCP *:8880 (LISTEN)

sw-cp-ser 11130 sw-cp-server 11u IPv6 195860 0t0 TCP *:8880 (LISTEN)

sw-cp-ser 11130 sw-cp-server 16u IPv4 238709 0t0 TCP h2191422.stratoserver.net:8880->pD9F084D8.dip0.t-ipconnect.de:52431 (ESTABLISHED)

root@h2191422:~# netstat -tulpn | grep 8443

tcp 0 0 0.0.0.0:8443 0.0.0.0:* LISTEN 11129/sw-cp-serverd

tcp6 0 0 :::8443 :::* LISTEN 11129/sw-cp-serverd

root@h2191422:~# netstat -tulpn | grep 8880

 

[maarten999]

And even the repair utility works again, which found no errors btw.

 

[UFHH01]

Hi maarten999,

In addition, I would like to mention, that your apache, nginx AND your PHP - versions and - handlers have absolute nothing to do with the Plesk Control Panel itself, which uses its very own webserver ( sw-cp-server ) and its very own PHP-FPM ( sw-engine ) - pls. consider to keep that in mind.

Pls. make as well sure to run

plesk installer install-all-updates​

... to avoid possible issues/errors/problems due to a possible "unfinished" Plesk - update/upgrade/patch -process.

 

[maarten999]

Installer still seems to have problems. Stalling at different point every time I restart. And within plesk it seems to have trouble connecting to the update server on port 8447