Virgin Reinstall - now can't manually approve newly registered users?!?!

[DaffyDuck]

When users get their registration emails, the link goes to a page that never finishes loading and generates a MYSQL error.

Likewise, when I attempt to manually approve new users, I get the 'LOADING' message at the bottom, and after approx a minute, the same 'TEMPORARY ERROR, MYSQL WENT AWAY' . so somehow, MYSQL appears to time out. So far everything else works, just not the new user registration / approval / verification.

 

Any ideas?

[DaffyDuck]

In fact, as an additional information, the same happens when I attempt to manually create a new user. 

This is on an UBUNTU server, with MYSQL and PHP 8.0 - the dashboard claims that everything is fine (ie all software versions are what it wants), yet this one glaring issue persists.

[Marc]

You would need to contact your hosting company regarding what is timing out there. The issue is unfortunately a hosting issue

[DaffyDuck]

I am self-hosted, and I'm my own hosting company. This is a virgin UBUNTU 22.04 installation on a VM running on an ESXI server.

Why does everything else work just fine, JUST the member approval / confirmation portion is failing?

 

[Jim M]

19 minutes ago, DaffyDuck said:

I am self-hosted, and I'm my own hosting company. This is a virgin UBUNTU 22.04 installation on a VM running on an ESXI server.

Why does everything else work just fine, JUST the member approval / confirmation portion is failing?

 

You would need to investigate your server to find the reason why your MySQL server is unavailable, I’m afraid. 

[DaffyDuck]

tail /var/log/nginx/chicabow.com.error.log

  thrown in /var/www/chicabow.com/htdocs/system/Db/Db.php on line 638" while reading response header from upstream, client: 104.28.123.87, server: chicabow.com, request: "POST /admin/?app=core&module=members&controller=members&do=add HTTP/2.0", upstream: "fastcgi://unix:/var/run/php/php80-two-fpm.sock:", host: "chicabow.com", referrer: "https://chicabow.com/admin/?app=core&module=members&controller=members"

2023/08/07 23:56:17 [error] 58345#58345: *512 FastCGI sent in stderr: "PHP message: PHP Fatal error:  Uncaught IPS\Db\Exception: MySQL server has gone away in /var/www/chicabow.com/htdocs/system/Db/Db.php:638

Stack trace:

#0 /var/www/chicabow.com/htdocs/system/Db/Db.php(972): IPS\_Db->preparedQuery()

#1 /var/www/chicabow.com/htdocs/system/Session/Admin.php(112): IPS\_Db->replace()

#2 [internal function]: IPS\Session\_Admin->write()

#3 /var/www/chicabow.com/htdocs/system/Redis/Redis.php(202): session_write_close()

#4 [internal function]: IPS\_Redis->IPS\{closure}()

#5 {main}

  thrown in /var/www/chicabow.com/htdocs/system/Db/Db.php on line 638" while reading response header from upstream, client: 104.28.123.87, server: chicabow.com, request: "POST /admin/?app=core&module=members&controller=members&do=add HTTP/2.0", upstream: "fastcgi://unix:/var/run/php/php80-two-fpm.sock:", host: "chicabow.com", referrer: "https://chicabow.com/admin/?app=core&module=members&controller=members"

This is part of the Invasion software:

MySQL server has gone away in /var/www/chicabow.com/htdocs/system/Db/Db.php:638

The line it is referencing, unfortunately, is just the part of the code that generates the error message. So, not helpful.

 

 

 

 

[Jim M]

1 hour ago, DaffyDuck said:

MySQL server has gone away

The error above is coming directly from your MySQL server. Our software is merely relaying that error. 

[Elon Report]

Daffy, 

Try creating a php file that pings the server after an increasing delay.   1 sec, then 2 sec, etc.   If you are seeing the error sooner then 8 hours, you have configured something to close the connection sooner.  

This page lists possible variables:

https://haydenjames.io/mysql-server-has-gone-away-error-solutions/