Search results

User name: burnley TITLE Obsidian on CentOS 7: Deleting a subscription with several domains attached via Plesk API killed Dovecot PRODUCT, VERSION, OPERATING SYSTEM, ARCHITECTURE OS CentOS Linux 7.7.1908 (Core) Product Plesk Obsidian Version 18.0.25 Update #2, last updated on April 8...

@abashurov thanks for looking into it, it's important for us to have it fixed quickly, since we're trying to keep our servers up to date to the best of our abilities. One option would be for you to ship your own ImageMagick5 compat package that installs somewhere in /opt/plesk and rebuild the...

User name: burnley TITLE Heads-up: RHEL7.8 is out, plesk-php-imagick should be rebuilt against libMagickCore-6.Q16.so.6 PRODUCT, VERSION, OPERATING SYSTEM, ARCHITECTURE CentOS 7.7 64bit, Plesk Obsidian, all plesk-php*-imagick packages PROBLEM DESCRIPTION With CentOS 7.8 around the...

User name: burnley TITLE Obsidian on CentOS 7: Plesk generates broken Dovecot configuration if client tries to secure email with invalid ssl certificate PRODUCT, VERSION, OPERATING SYSTEM, ARCHITECTURE 18.0.23 CentOS 7 1800200130.12 PROBLEM DESCRIPTION We had a client today...

@IgorG yeah, I knew there was a place to report bugs, just couldn't find it yesterday, it's "burried" in Onyx section. Could you perhaps create a similar sticky for Obsidian?

Stay safe everyone!

Sorry for cross-posting, but it's causing server-wide POP3/IMAP service outage. The original post is here: Issue - Obsidian on CentOS 7: Plesk generates broken Dovecot configuration if client tries to secure email with invalid ssl certificate Can you please look at it asap and push a fix? Or at...

18.0.23 CentOS 7 1800200130.12 We had a client today fiddling with the SSL/TLS Certificates feature and they ended up with an entry that contain the CSR and key components. No certificate and no CA certificate. However that didn't prevent the client to [ab]use "Secure Email" option, which...

You have been assimilated by Plesk. We are Plesk. Resistance is futile! hot tip: The next version of Plesk will be called "Cube" (-;

Posted my question here as well, since it's become a major nuisance for us and we want it fixed: [FIXED BUG] Let's Encrypt challenge files are not removed from .well-known/acme-challenge/ directories Just stumbled on this article after a major clean up on our Linux plesk servers, littered with...

That's right @IgorG we're running higher apr version on this test server, so not an issue: rpm -q apr apr-1.6.5-1.x86_64 In live we're running apr-1.4.8-5.el7.x86_64

Great, thanks for confirming the mod_fcgid changes, it now all makes sense. On the mod_security version mismatch, I'm using the rpm shipped by Plesk: [root@plesk12 ~]# rpm -qa | grep mod_security mod_security-2.9.3-2.centos.7+p18.0.23.3+t200130.1239.x86_64 [root@plesk12 ~]# updatedb...

@IgorG I've updated the issue: https://talk.plesk.com/threads/after-onyx-obsidian-upgrade-we-see-graceful-req-false-followed-by-stop-start-true-in-apache_restart_stat-log.355286/post-872865 Suddenly, after the upgrade, mod_fcgid behaviour has changed and now "httpd -k graceful" doesn't work...

Think I might have found the reason. It's not in Plesk's cli tool, it's in the systemd unit file using for httpd. Because of the Apache semaphores leak issue we were using a custom httpd.service file that would remove the Apache semaphores on service reload as a workaround. However, after...

I think there might be an issue with /usr/local/psa/admin/sbin/apache_control_adapter tool. On an affected test server I see this: /usr/local/psa/admin/sbin/apache_control_adapter --restart --http-port 7080 --https-port 7081 --graceful -d apache ready test: request with 40 sec timeout finished...

After Onyx -> Obsidian upgrade we see "graceful-req",false followed by "stop-start",true in apache_restart_stat.log

User name: burnley TITLE After Onyx -> Obsidian upgrade we see "graceful-req",false followed by "stop-start",true in apache_restart_stat.log PRODUCT, VERSION, OPERATING SYSTEM, ARCHITECTURE ‪Plesk Obsidian Version 18.0.23 Update #3, last updated on Feb 10, 2020 04:37 PM CentOS Linux...

@IgorG anyone looking into this issue for us? I've enabled debugging in panel.ini and here's what I see when reconfiguring the php handler on a subscription: /var/log/plesk/panel.log [...] [2020-02-11 14:43:01.969] DEBUG [panel] Service_Driver_Web_Server_Apache::logRestartForStatistics...

@IgorG I think there are 2 separate issues here, both related to reloading httpd: 1. The graceful-req followed by stop-start actions. We weren't seeing this pattern in Onyx. Is this behaviour normal and expected? 2. The 502 errors on graceful, which appear to be caused by mod_fcgid. Before the...

@IgorG the problem here are not the 502, but the graceful-req followed by stop-start actions. The 502 & 503 could be the effects.