Issue 502 Bad Gateway after I changed one of domain names

[typhone]

Server operating system version

ubuntu 18

Plesk version and microupdate number

Version 18.0.52 Update #3

Hi,
yesterday I got an error on all websites hosted, 502 Bad Gateway.
I did rename one domaine and after that I start getting some errors on web domain page.




what should I do?

 

[Peter Debik]

If removal/recreation operations are done too fast, sometimes race conditions occur that can lead to false positive error messages. It then can help to simply reconfigure the affected domain configuration files.

In Web Configurations Trouble Shooter, please try to repair the "Broken" config entries only.



Please report here if that solved the issue.

Do you have custom webserver configuration templates or are you using the default?

 

[typhone]

the problem now is I tried to restart the server from plesk panel, and now everything is down, the server in my cloud panel is up but plesk panel is down and all website linked to it,
I tried to restart apache using ssh but I got this error

 

[Peter Debik]

Do you mean with "now" that after the repair transaction the server is down?

Plesk runs on its own web server. It is not affected by the previously recommended action. Can you please check whether the services sw-cp-server and sw-engine are both running? (service sw-cp-server status, service sw-engine status)?

 

[typhone]

When the errors mentioned above first occurred, I tried using repair tools. However, I consistently received an error stating that the directory was not found. Afterward, I attempted to restart the server from the Plesk panel, but now I am unable to access the Plesk panel.

 

[typhone]

I tried your solution but I got these errors
any idea?

 

[Peter Debik]

We're talking about two different issues here.
1) You cannot access Plesk
2) Web server configuration files cannot be created

I assume that the screenshot from Web Server Configurations Trouble Shooter was taken before the issue with not being able to access Plesk occured.

1) Please check that the disk isn't full, e.g. with df.

2) Please check whether the directory /var/www/vhosts/system exists. Several web domain directories that should descende from that seem to be missing. Please verify that /var/www/vhosts/system/<domainname> exist for other domains. Did you maybe set a custom hostname or is the domain that you are using for websites also used as the hostname? Were any other changes done to the paths where websites are stored, e.g. did you try to change the /var/www/vhosts path to something else before?[/icode]

 

[typhone]

now I can access to my plesk.
I tried to check var but I found just tmp folder

 

[typhone]

I run ls command here is the results

 

[typhone]

No, I didn't change anything in path. I simply renamed 'aurorafast' to 'aurorafastt,' in hosting settings and then this issue occurred.

 

[Peter Debik]

There is a possibility that you overlooked a red error message when renaming. Frequently renaming cannot be completed due to running processes. If that happens, only part of the transactions that are needed for renaming a domain where completed.

For that reason, when you rename an account, first "Deactivate" it (not remove, not suspend, but "deactivate" from the domain drop down of the main domain), then rename the domain. After successful renaming, "activate" it again. This will ensure that running processes are stopped and all necessary parts can be renamed. What I see in your output above is that the system configuration directories of the domain where not renamed.

Then, if it was my server, I'd use the above mentioned procedure to rename the domain back to what is was named before and check whether the error could be solved by that. If not, try plesk repair web on the Linux console or let Web Server Configurations Troubleshooter reconfigure the domain. If that completes successful use the above mentioned procedure to rename the domain again to what you actually want.

I'd not attempt solving this by renaming the /var/www/vhosts/system/* directories, because service configuration files beyond Apache, such as Nginx or PHP or mail server files, may have been updated properly, and these might point to wrong filenames after such a manual attempt.

 

[typhone]

I tried web rebuild broken and select but those files are stying the same. do you think deleting them is a good idea?

 

[Peter Debik]

The point was to rename the domains back to what they were named. Was that done prior to repair?

 

[WebHostingAce]

I didn't read the whole thread.

For you last screenshot, try to turn off the ModSecurity and rebuild the configuration. Then turn it back on.

I experieced the same issue after this https://support.plesk.com/hc/en-us/...26-2023-client-denied-by-server-configuration Atomic Update but this might not be related to you.