• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

HELP. Control panel white page and psa mysql connections filling up.

Bjorn

Basic Pleskian
Hi,

Operating system: Centos 7.1.1503
Current used Plesk- Product: 12.5.30 MU #23
Depending installed mail - server software: Postfix

Currently i'm experiencing a major problem with my Plesk install. When I try to access the control panel I get a blank page which (after around 10 min) switches to a '504 Gateway Time-out nginx' message. I tried port 8443 and 8880.

Also the mysql service (mariadb) stops working because there are too many connections. Obviously this breaks ALL sites on the server.....!

When I restart maraidb and check the mysqladmin processlist I see its filling up with:

Id:62 | User:Admin | Host:localhost | db: psa | Command:Sleep | Time:3 | State: | Info: | Progress:0.000

To me, it looks like the psa db is corrupted...
I checked all the logs I could find, but nothing about a psa db lookup error.

Tail log files
[var/log/sw-cp-server/sw-engine.log]
NOTICE: fpm is running, pid 20301
NOTICE: ready to handle connections
NOTICE: systemd monitor interval set to 10000ms

[var/log/sw-cp-server/error_log]
[error] 1507#0: *68 upstream timed out (110: Connection timed out) while reading response header from upstream, client: [my personal ip], server: , request: "GET /login.php3?success_redirect_url= HTTP/1.1", upstream: "fastcgi://unix:/var/run/sw-engine.sock",
[usr/local/psa/admin/logs/panel.log]
Here I see the GET log for trying to access the control panel. Not an error message.

I really need some help fast :(, currently i can't go to sleep because i need to restart mysql when the mysql connections reaches max. I allready set it to max_connections=300.

Also, a few weeks ago i noticed i couldn't create a subdomain for a particular domain. All other domains worked fine.

Any help will be much appreciated!

Regards, Bjorn
 
Sir, YOU ARE AWESOME :)

I can access Plesk, and the mysqladmin processlist remains clean now!

Also, I had some problems with domain / customer creation (plesk hanged), that also seems to work fine now after the first tests... WOW NICE..

I maybe found that kb article myself, but probably didn't do anything with it because it states 'After update'. I did do an update, but that was +/- 10 days ago.

You really made my day UFHH01, thanks for that!
 
Follow up question.

Is this something I have to do on a regular basis? This because the KB states it's a known software issue and will be fixed in future updates?

And it would also be great if you can give me some insight on why/where this problem occurs. I understand it has something to do with the cache.

Bjorn
 
And 1 more thing about the KB article.

Step 2. plesk installer install-all-updates
I get the following message:
Doing restart of Plesk Installer

BUSY: Update operation was locked by another update process.
exit status 1


Should i do anything with this?

[EDIT]

There was another update process running that was timed out.
I killed the pid and started the update process which completed OK :)
 
Last edited:
Is this something I have to do on a regular basis? This because the KB states it's a known software issue and will be fixed in future updates?

"normally" not.... so why my quotes? Well... we don't know, when the bug is completely solved, so you might bookmark it for now and you will notice on this specific page, when the bug has been eliminated and a patch has been released ( Plesk-Team-Member update the page as soon as a patch is out ).
 
Ok, will do that. Thanks again for helping me, all seems up and running great again!
 
Hi,
After updating Plesk i can confirm the software issue #PPPM-3485 is not yet fixed.

Please note that after the Plesk update i have to clear the cache again (step 1: kb article).

Regards,
Bjorn
 
Back
Top