Input Plesk 17.8 good with CentOS 7.7

[Bitpalast]

Just to let you know that we've updated all systems to CentOS 7.7 and are not yet seeing any issues with Plesk on it.

 

[Sergio Manzi]

Hello @Peter Debik !

I've just updated too (I did that through the Plesk control panel "System Updates") and so far everything seems to be OK.
Anyway while updating I received an error message and I'm wondering if this should be of any concern:

Internal error ;-P

ERROR: Zend_Db_Adapter_Exception: SQLSTATE[HY000] [2002] No such file or directory (Abstract.php:144)<br />
<br><a href='http://kb.plesk.com/plesk-error/search?metaId=1751d5edcca5340f78144a4499e8c1d5&messageId=2eb8d2699e722503bf79ae1c7819dc4e&file=Abstract.php&line=144&type=Zend_Db_Adapter_Exception&version=17.8.11&message=SQLSTATE%5BHY%5D+%5B%5D+No+such+file+or+directory' target='_blank'>Search for related Knowledge Base articles</a>

Plesk is (and was) Version 17.8.11 Update #67

Cheers!

 

[Bitpalast]

Absolutely no problem to have this message. When daemons restart, there is a short interruption of their service on the system. Zend Framework can then not communicate with them which results in the error message. A short moment later, when the new versions of the services (or binaries or whatever) are running again, the framework can connect again and the error message is obsolete. So this type of error message is more or less by design, but no real problem.

 

[Sergio Manzi]

Thank-you very much, Peter!

 

[Kingsley]

So you are using new plesk now?

 

[Ales]

We just had an issue, so I'm putting this here in case anyone else stumbles upon it. Note, this only happens in a specific OS setup and if an outdated kernel is used.

Some outdated kernels, such as OpenVZ 6 kernels prior to vzkernel-2.6.32-042stab134.7, cause certain services within CentOS 7.7 virtual machines and systemd-67.el7_7.1 to fail.

Updated systemd treats symlinks differently, so, in combination with the outdated kernel, systemd unit files with references to pid files within /var/run (which is a symlink to /run) fail to start.

Errors include "Can't open PID file /var/run/spamd.pid (yet?) after start: Too many levels of symbolic links" in case of Spamassassin and others, sometimes not so obvious ones, for other services.

Commands that might help you diagnose the issue are:

uname -a
systemctl list-units --all
systemctl status fail2ban.service
systemctl status monit.service
systemctl status sshd.service
systemctl status spamassassin.service
systemctl status xinetd.service

The proper solution is to update the kernel. In case of OpenVZ 6, using a more recent, preferably latest vzkernel, should be fine.

For those that can't update the kernel for some reason, the workaround is to adjust the service file by overriding the relevant part, e.g. for Spamassassin:

systemctl edit spamassassin.service

enter the following content:

[Service]

PIDFile=/run/spamd.pid

Reload systemd and restart the service:

systemctl daemon-reload
systemctl restart spamassassin.service

 

[themew]

Did the update today via SSH... fast downloading and replacing of files but the server hung after CLEANUP was finished. So, I exited the server, and logged back in. YUM UPDATE showed no files to update so rebooted and logged into the back of Plesk after several minutes.

No issues to report and the server is serving pages about 100ms faster than before. Just posting for other CentOS updaters.