• Please be aware: Kaspersky Anti-Virus has been deprecated
    With the upgrade to Plesk Obsidian 18.0.64, "Kaspersky Anti-Virus for Servers" will be automatically removed from the servers it is installed on. We recommend that you migrate to Sophos Anti-Virus for Servers.
  • The Horde webmail has been deprecated. Its complete removal is scheduled for April 2025. For details and recommended actions, see the Feature and Deprecation Plan.
  • We’re working on enhancing the Monitoring feature in Plesk, and we could really use your expertise! If you’re open to sharing your experiences with server and website monitoring or providing feedback, we’d love to have a one-hour online meeting with you.

Resolved 17.5.3 > 17.8.11 Upgrade - Issues

learning_curve

Golden Pleskian
We've run the CLI 17.5.3 > 17.8.11 upgrade process a couple of times, which went quite well, but we're still on 17.5.3 when posting this thread. There were some issues (shown below) the biggest being that sw-cp-server won't run after the upgrade. We run a cloud server, so if there's impactive upgrade issues, we immediately revert back to the pre-upgrade server snapshot with just a few seconds downtime / almost negligable loss of web service. This means we can go and investigate the upgrade issues (e.g. this thread...) solve them and then run the the CLI 17.5.3 > 17.8.11 upgrade process again, sucessfully, if we're lucky ;)

1) At the pre-upgrade check stage, we get the same error that's already been posted in other threads, relating to web server templates located in /usr/local/psa/admin/conf/templates/custom/ However, like other users have posted, we have no custom templates located in there anyway... and, if we run a previoulsy posted solution, which was; run # plesk repair db then # plesk repair web before the upgrade and then try running the upgrade again We still get exactly the same error message (warning?) at the pre-upgrade check stage.

2) At the changing shell for mhandlers-user stage, we get warnings about missing files e.g.
Code:
Cleanup: plesk-roundcube [78/135]
warning: file /var/tmp/plesk-roundcube: remove failed: No such file or directory
3) At the same stage, we also get errors which appear at this (early) stage, to be of no immediate detrimental effect e.g.
Code:
Updating: psa-selinux-17.8.11-cos7.build1708180212.17.noarch [70/135]
Failed to resolve roletype statement at /etc/selinux/tmp/tmp/modules/400/plesk/cil:2
semodule:  Failed!
Failed to resolve roletype statement at /etc/selinux/final/tmp/modules/400/plesk/cil:2
semodule:  Failed!
4) During the Installing Admin Server stage, we get this failure / warning and it is repeated, at the end of this stage too:
Code:
which: no unrar in (/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin)
Trying to install sftp-server binary into chroot environment... cp: cannot stat 'internal-sftp': No such file or directory
done
/var/www/vhosts/chrootinternal-sftp: inode/directory; charset=binary
probably it will not work in chrooted accounts
WARNING!
Some problems are found during register /var/www/vhosts/chrootinternal-sftp in chrooted environment(see log file: /var/log/plesk/install/plesk_17.8.11_installation.log)
5) Then at the end, we see:
Code:
Starting sw_cp_server service... failed
Followed soon after by
Code:
...
All files were downloaded.
All patches were applied.                                                                                                                
Patches were installed successfully.
Launching component and product checkers...
Downloading file PSA_17.8.11/examiners/congratulations.sh: 100% was finished.
Components and product check results:
Upgrade is finished
Congratulations...!
But ;) having seen earlier that sw_cp_server service failed to start message, we ran
Code:
systemctl status sw-cp-server.service
Which tells us
Code:
● sw-cp-server.service - Startup script for Plesk control panel server
Loaded: loaded (/usr/lib/systemd/system/sw-cp-server.service; enabled; vendor preset: disabled)
Active: failed (Result: exit-code) since Mon 2018-07-16 03:18:30 BST; 40s ago
Process: 29842 ExecStartPre=/usr/sbin/sw-cp-serverd -q -t (code=exited, status=1/FAILURE)
Main PID: 3665 (code=exited, status=0/SUCCESS)
And if we investigate further, we see
Code:
* systemd[1]: Starting Startup script for Plesk control panel server...
* sw-cp-serverd[29842]: nginx: [emerg] unknown directive "auth_request" in /etc/sw-cp-server/conf.d/repair_plesk.inc:5
* sw-cp-serverd[29842]: nginx: configuration file /etc/sw-cp-server/config test failed
* systemd[1]: sw-cp-server.service: control process exited, code=exited status=1
* systemd[1]: Failed to start Startup script for Plesk control panel server.
* systemd[1]: Unit sw-cp-server.service entered failed state.
* systemd[1]: sw-cp-server.service failed
We've looked, but can't see a reference to that particular error anywhere else. There was no file named: /etc/sw-cp-server/conf.d/repair_plesk in the /etc/sw-cp-server/conf.d folder with which, we could try and trace any possible reasons for this error... o_O It appears that it may possibly be related to permissions, but, where, we're not sure at this stage. FWIW the file /etc/sw-cp-server/config itself, plus the contents of /etc/sw-cp-server/ and /etc/sw-cp-server/conf.d all seemed to be unchanged...

We think 2) is a set of standard warnings and isn't an issue for us
We think 3) should be okay when we work further on selinux and therefore isn't an issue for us
We think 4) is fine for us because we don't use or have, any chrooted domains anyway (currently)

We'd like to fix 1) and obviously, we do need to fix 5) because that prevented us from even logging into the 17.8.11 Plesk Panel! Anyone from Plesk, your help would be most welcome
 
sw-cp-serverd[29842]: nginx: [emerg] unknown directive "auth_request" in /etc/sw-cp-server/conf.d/repair_plesk.inc:5
It looks very strange. I have not found any mentions of this error in all our internal resources.
What is your OS? Maybe you have some kind of customizations on your server?
 
It looks very strange. I have not found any mentions of this error in all our internal resources. What is your OS? Maybe you have some kind of customizations on your server?
Our OS is shown on our forum signature (CentOS 7.5.1804) but the issue is elsewhere we think? The fact that you @IgorG cannot find any identical error message elsewhere and after we said, that we couldn't either, makes us think this is releated to a customization. We do have several customizations, including cw-sp-server itself :p which currently, we have on our server twice (see below). Based on this, we'll remove our customized cw-sp-server and then try the upgrade again. As mentioned, we have the ability to server snapshot back after trying this (if needed) which is handy.

Anything new on issue 1)?

FWIW The reasons why we customized sp-cw-server are covered in other forum threads. We've had zero problems with this customized cw-sp-server during ALL the 17.5.3 upgrades to date. However, this latest upgrade is 17.5.3 > 17.8.11 so... there maybe some subtle Plesk difference now that causes this new issue.

sw-cp-server > We have the standard Plesk version, which can be upgraded without any problems via plesk and/or yum update. Then we have our own customized verison too, which won't ever get upgraded and which we are intending to remove anyway (once TLSv1.3 is officially released and an updated cw-sp-server with TLSv1.3 is then released by Plesk soon after)

Here's the data from the two cp-sw-server installations:
Code:
[root@root]# rpm -qi sw-cp-server

Name        : sw-cp-server
Version     : 2.4.0
Release     : centos7.17022818
Architecture: x86_64
Install Date: Mon 24 Apr 2017 05:58:43 BST
Group       : System Environment/Daemons
Size        : 1231011
License     : 2-clause BSD-like license
Signature   : DSA/SHA1, Fri 17 Mar 2017 09:36:13 GMT, Key ID bd11a6aa914bdf7e
Source RPM  : sw-cp-server-2.4.0-centos7.17022818.src.rpm
Build Date  : Tue 28 Feb 2017 11:18:58 GMT
Build Host  : bcos7x64.plesk.ru
Relocations : (not relocatable)
Packager    : Plesk <[email protected]>
Vendor      : Plesk
Summary     : Plesk control panel server
Description :
HTTP server for Plesk control panel

Name        : sw-cp-server
Version     : 2.6
Release     : 1
Architecture: x86_64
Install Date: Sat 09 Sep 2017 00:42:53 BST
Group       : checkinstall
Size        : 8594294
License     : GPL
Signature   : (none)
Source RPM  : sw-cp-server-2.6-1.src.rpm
Build Date  : Sat 09 Sep 2017 00:38:51 BST
Build Host  : ***our own server name***
Relocations : (not relocatable)
Packager    : checkinstall-1.6.2
Summary     : Server For Plesk Control Panel
Description :
Server For Plesk Control Panel
 
As posted on another thread (different topic but related in other ways) we've solved this by taking a new cloud server, with a clean install of Ubuntu 18.04.01 and a clean install of Plesk 17.8.11#16 and then subsequent additional mods. We're 'swapping' everything over from our CentOS7 / Plesk 17.5.3 cloud sever on to its Ubuntu replacement, testing everything, taking the new Ubunto server live and then deleting the old CentOS7 server. There were a couple of minor issues, when running a clean install of Plesk 17.8.11#16 on Ubuntu 18.04.01 but we'll post those separately & they were not related to this thread's ^^ issue anyway.
 
Back
Top