Question Small Plesk Errors in the logs - Config Related - Local or Plesk or OS?

[learning_curve]

Server operating system version

Ubuntu 24.04.1 LTS

Plesk version and microupdate number

18.0.65 #1

As per the the thread title. No visible traces of any of these and no performance penalties observed, but obviously, there's a mis-config (or two) somewhere.
Anybody whose had the same, but resolved them and / or anybody from Plesk care to comment?

The first one appears to be panel related, but only appears (in the logs) after server re-boots:

DB query failed: SQLSTATE[HY000] [2002] No such file or directory:
0: /opt/psa/admin/plib/Db/Adapter/Pdo/Mysql.php:79
Db_Adapter_Pdo_Mysql->query(string 'SET sql_mode = ''')
1: /opt/psa/admin/plib/CommonPanel/Application/Abstract.php:113
CommonPanel_Application_Abstract::initDbAdapter()
2: /opt/psa/admin/plib/Session/Helper.php:176
Plesk\Session\Helper::initStorage()
3: /opt/psa/admin/plib/auth.php:275
AutoPrepend->initUserSession()
4: /opt/psa/admin/plib/auth.php:188
AutoPrepend->run()
5: /opt/psa/admin/plib/auth.php:582

The second one, again appears to be panel related (FileManager) but it's daily, so cron generated (probably):

'/opt/psa/admin/bin/filemng' 'root' 'exec' '/' 'bash' '-c' 'timeout 5 /usr/bin/pro status --format json' '--allow-root' failed with code 124.
stdout:
stderr:

The third is directly related to the second, but is util_exec derived i.e. access to in this case.

proc_close() failed ['/opt/psa/admin/bin/filemng' 'root' 'exec' '/' 'bash' '-c' 'timeout 5 /usr/bin/pro status --format json' '--allow-root'] with exit code [124]

We have many things buttoned right down, in terms of access / rights etc hence these errors maybe collateral damage from our own config somehwere else, unrelated.
The fact that there's no visible evidence (other than these log entries) could easily make this a diminishing returns exercise, hence this post, just in case there's a short cut!
FWIW - The last two only started recently after updates for both Plesk and Ubuntu.

 

[learning_curve]

~~

'/opt/psa/admin/bin/filemng' 'root' 'exec' '/' 'bash' '-c' 'timeout 5 /usr/bin/pro status --format json' '--allow-root' failed with code 124.
stdout:
stderr:

~~

proc_close() failed ['/opt/psa/admin/bin/filemng' 'root' 'exec' '/' 'bash' '-c' 'timeout 5 /usr/bin/pro status --format json' '--allow-root'] with exit code [124]

These last 2 ^ are related to Ubuntu Advantage (as it was named previously) or Ubuntu Pro (as it is named now). Not sure that filemanager is anything but a messenger here, just relaying the issue and then its conclusion with a timeout code 124, that then prevents any further efforts. Utilising cli, it's really quite simple to: # /usr/bin/pro status --format json when SSH'd in as root, although usr/bin/pro is just a symlink to /usr/bin/ubuntu-advantage anyway & this can also be # /usr/bin/ubuntu-advantage status --format json when SSH'd in as root. Both, provide exactly the same values, without errors or warnings. There's no access rights or group permissions that could / would prevent access to any of the services or checks related to Ubuntu Advantage / Ubuntu Pro, so the current supposition, is that there's a variable (but one that can be altered) within: /lib/systemd/system/ua-reboot-cmds.service that's causing these errors/warnings, as these 2, are also only appearing in the logs now, after a specific time following re-boots. Not looked at the 1st one yet, but will do, pretty soon.

 

[FISHeR]

Hi. Thanks for reporting the problems. The last 2 errors are related. The bug ID is PPP-66830. Probably will be fixed in the upcoming version.

 

[theunknownstuntman]

As per the the thread title. No visible traces of any of these and no performance penalties observed, but obviously, there's a mis-config (or two) somewhere.
Anybody whose had the same, but resolved them and / or anybody from Plesk care to comment?

The first one appears to be panel related, but only appears (in the logs) after server re-boots:

DB query failed: SQLSTATE[HY000] [2002] No such file or directory:
0: /opt/psa/admin/plib/Db/Adapter/Pdo/Mysql.php:79
Db_Adapter_Pdo_Mysql->query(string 'SET sql_mode = ''')
1: /opt/psa/admin/plib/CommonPanel/Application/Abstract.php:113
CommonPanel_Application_Abstract::initDbAdapter()
2: /opt/psa/admin/plib/Session/Helper.php:176
Plesk\Session\Helper::initStorage()
3: /opt/psa/admin/plib/auth.php:275
AutoPrepend->initUserSession()
4: /opt/psa/admin/plib/auth.php:188
AutoPrepend->run()
5: /opt/psa/admin/plib/auth.php:582

The fact that there's no visible evidence (other than these log entries) could easily make this a diminishing returns exercise, hence this post, just in case there's a short cut!
FWIW - The last two only started recently after updates for both Plesk and Ubuntu.

I happen to have the same error, after upgrading from Ubuntu 18.0x LTS to 20.0x LTS. Plesk has not been upgraded as part of the Ubuntu upgrade.
I see this error in Plesks Logbrowser, for "Panel". It happens after I login to Plesk, which works fine, or when I restart the mysql.service.

Plesk "check & repair" does not report any issues.

Does anyone have an idea? The several threads here in this forum did not help. Also The error in Plesk: ERR [panel] SQLSTATE[HY000] [2002] No such file or directory, Can't connect to local MySQL server through socket was no help.

Product version: Plesk Obsidian 18.0.73.3
OS version: Ubuntu 20.04 x86_64
Build date: 2025/10/09 17:00
Revision: 02428526e4b72b7901f6fba310b46fb9d07c2990

DB query failed: (2002) SQLSTATE[HY000] [2002] No such file or directory:
0: /opt/psa/admin/plib/Db/Adapter/Pdo/Mysql.php:79
Db_Adapter_Pdo_Mysql->query(string 'SET sql_mode = ''')
1: /opt/psa/admin/plib/CommonPanel/Application/Abstract.php:113
CommonPanel_Application_Abstract::initDbAdapter()
2: /opt/psa/admin/plib/Session/Helper.php:176
Plesk\Session\Helper::initStorage()
3: /opt/psa/admin/plib/auth.php:269
AutoPrepend->initUserSession()
4: /opt/psa/admin/plib/auth.php:175
AutoPrepend->run()
5: /opt/psa/admin/plib/auth.php:576

 

[theunknownstuntman]

Regarding The error in Plesk: ERR [panel] SQLSTATE[HY000] [2002] No such file or directory, Can't connect to local MySQL server through socket
it is strange, that I can not find any reference to "mysql.sock" in any mentioned files, that are mentioned in the "resolution" section.

There is a "mysqld.sock" and "mysqlx.sock" in "/var/run/mysqld/" though.

I appreciate any help in this case. I would love to upgrade to Ubuntu 22.0x LTS in the next step, but don't want to start, if I am not 100% certain, that everything works as expected.

 

[learning_curve]

~~ I happen to have the same error, after upgrading from Ubuntu 18.0x LTS to 20.0x LTS. Plesk has not been upgraded as part of the Ubuntu upgrade.
I see this error in Plesks Logbrowser, for "Panel". It happens after I login to Plesk, which works fine, or when I restart the mysql.service ~~

FWIW Plesk themselves, should have addressed this error some time ago. They did indeed, fix the other two issues (see posts #2 and #3 above) so it was a mystery why they just appeared to ignore this one. However, FWIW we're using the current release of Ubuntu and for whatever reason (pure luck for Plesk maybe ) this issue is no longer present (for us).

You've already your upgraded Ubuntu OS once, but when using Plesk with the Ubuntu release that you're now using (and the next one) you might still continue to see the error. It's a non-problematical error in reality, as you've discovered, so it doesn't warrant a vast amount of time tracing and bug fixing it. Unless it appears on a Plesk fix-list soon, you might just be stuck with it, until you choose to upgrade to the current Ubuntu OS release in the future.

~~ Regarding The error in Plesk: ERR [panel] SQLSTATE[HY000] [2002] No such file or directory, Can't connect to local MySQL server through socket
it is strange, that I can not find any reference to "mysql.sock" in any mentioned files, that are mentioned in the "resolution" section.
There is a "mysqld.sock" and "mysqlx.sock" in "/var/run/mysqld/" though ~~

Yes. that's an old fix now, which we didn't use and which is why it is perhaps not relevant to this specific issue anymore (depending on your OS and Config etc).
One simple example: A few MariaDB releases ago, (as part of the big project of removing all references to MySQL inside MariaDB) It was advisable to use mariadb.cnf NOT my.cny. This, plus many other socket and database changes too.

~~ I appreciate any help in this case. I would love to upgrade to Ubuntu 22.0x LTS in the next step, but don't want to start, if I am not 100% certain, that everything works as expected.

Not really helped you a lot here, however the error shown in your first post, shouldn't stop you running a succesful upgrade from Ubuntu 20.* to Ubuntu 22.*
Many other things might... but that specific error, despite it still producing the error reports, shoud have no effect on the upgrade process itself

 

[Sebahat.hadzhi]

The error in question indicates that Plesk was unable to connect to the MySQL service. As far as I am aware, this often happens if there's interruption of the database service. Could you please check the status systemctl status mariadb and confirm if there is any correlation between the running time and appearance of the error?

 

[learning_curve]

The error in question indicates that Plesk was unable to connect to the MySQL service. As far as I am aware, this often happens if there's interruption of the database service. Could you please check the status systemctl status mariadb and confirm if there is any correlation between the running time and appearance of the error?

FWIW @Sebahat.hadzhi We were aware of this possible cause ^ but short inspections (post #1 - diminishing returns exercise...) determined these to more than likely, be false positives, because in our case, they only occurred after server re-boots (timing of the connect request versus Ubuntu boot process perhaps?). They don't occur anymore now (in our case) but seeing as both Plesk and Ubuntu have made may updates and upgrades since back then, we obviously can't identify the actual, true cause from back then anyway. It may be different for @theunknownstuntman

 

[theunknownstuntman]

FWIW Plesk themselves, should have addressed this error some time ago. They did indeed, fix the other two issues (see posts #2 and #3 above) so it was a mystery why they just appeared to ignore this one. However, FWIW we're using the current release of Ubuntu and for whatever reason (pure luck for Plesk maybe ) this issue is no longer present (for us).

You've already your upgraded Ubuntu OS once, but when using Plesk with the Ubuntu release that you're now using (and the next one) you might still continue to see the error. It's a non-problematical error in reality, as you've discovered, so it doesn't warrant a vast amount of time tracing and bug fixing it. Unless it appears on a Plesk fix-list soon, you might just be stuck with it, until you choose to upgrade to the current Ubuntu OS release in the future.

Yes. that's an old fix now, which we didn't use and which is why it is perhaps not relevant to this specific issue anymore (depending on your OS and Config etc).
One simple example: A few MariaDB releases ago, (as part of the big project of removing all references to MySQL inside MariaDB) It was advisable to use mariadb.cnf NOT my.cny. This, plus many other socket and database changes too.

Not really helped you a lot here, however the error shown in your first post, shouldn't stop you running a succesful upgrade from Ubuntu 20.* to Ubuntu 22.*
Many other things might... but that specific error, despite it still producing the error reports, shoud have no effect on the upgrade process itself

Thanks for your detailed answer. I decided to upgrade the server and with a few problems I was able to finish this. See below.

 

[theunknownstuntman]

The error in question indicates that Plesk was unable to connect to the MySQL service. As far as I am aware, this often happens if there's interruption of the database service. Could you please check the status systemctl status mariadb and confirm if there is any correlation between the running time and appearance of the error?

Thank you for your answe @Sebahat.hadzhi . At this time mysql.service was still running (mysql-server-8.0 to be precise). I decided after taking a snapshot for the server to upgrade to Ubuntu 22.0x LTS, as explained in your upgrade-guide.

I ran in the same error, as mentioned above, and was not able to login to Plesk. mysql.service was also not able to start ("Could not open file '/var/log/mysql/error.log'"). Reinstalling mysql-server-8.0 fixed that thankfully.

 

[theunknownstuntman]

Strange though "The error in Plesk: ERR [panel] SQLSTATE[HY000] [2002] No such file or directory, Can't connect to local MySQL server through socket" in the Log Browser regarding the "Plesk", "Panel", I started with is still visible...

 

[learning_curve]

Thank you for your answe @Sebahat.hadzhi . At this time mysql.service was still running (mysql-server-8.0 to be precise). I decided after taking a snapshot for the server to upgrade to Ubuntu 22.0x LTS, as explained in your upgrade-guide.

I ran in the same error, as mentioned above, and was not able to login to Plesk. mysql.service was also not able to start ("Could not open file '/var/log/mysql/error.log'"). Reinstalling mysql-server-8.0 fixed that thankfully.

Apologies @theunknownstuntman Ignore any MariaDB references made. Didn't realize that you were using MySQL not MariaDB until you made these latest posts, because you have no forum signature, plus many things are still labelled / referenced MySQL anyway, even though they are actually MariaDB.

The error still being there, in Ubuntu 22.* which you've successfully upgraded to now, we did predict, in post #6 above, but we did also say that (in our case) it no longer occurs now with the latest Ubuntu release c/w the latest Plesk release (as per our forum signature, at the time of writing this post).

FWIW The upgrade from Ubuntu 22.04.* LTS to Ubuntu 24.04.* LTS whilst using Plesk for hosting, on the same server(s), is a bit more challenging than the one you've just done. Read lots in advance, before commencing with this one.