Migration returns nginx errors

[tkalfaoglu]

Username: tkalfaoglu

TITLE

Migration returns nginx errors

PRODUCT, VERSION, OPERATING SYSTEM, ARCHITECTURE

centos 7 obsidian to centos 8 obsidian

PROBLEM DESCRIPTION

When migrating domains 40 at a time, I received 5 of these emails:

Unable to generate the web server configuration file on the host <ocean.kalfaoglu.net> because of the following errors:
Template_Exception: nginx: [emerg] open() "/etc/nginx/plesk.conf.d/vhosts/akdenizdoseme.com.conf" failed (2: No such file or directory) in /etc/nginx/conf.d/zz010_psa_nginx.conf:7
nginx: configuration file /etc/nginx/nginx.conf test failed
file: /usr/local/psa/admin/plib/Template/Writer/Webserver/Abstract.php
line: 75
code: 0
Please resolve the errors in web server configuration templates and generate the file again.

STEPS TO REPRODUCE

Select migration manager, select all domains with letter 'A' start migration..

ACTUAL RESULT

migrates, but gives some nginx errors

EXPECTED RESULT

Smooth sailing

ANY ADDITIONAL INFORMATION



YOUR EXPECTATIONS FROM PLESK SERVICE TEAM

Confirm bug

 

[IgorG]

Have you checked first that configuration of these domains on the source server is ok before starting the migration process?
Do you think the Plesk Migrator broke the configuration of these domains during the migration process?

 

[tkalfaoglu]

ocean is the destination server, so I assumed that the problem was with the destination?

 

[Monty]

I remember I had a similar issue a couple of years ago. In my case, Apache/nginx was reloaded too often/quickly while Plesk was still busy building the configuration files for the migrated domains.

The migration continued and the errors disappeared after all domains have been migrated.
@tkalfaoglu You may want to try this: On the destination server go to "Apache & nginx settings" and set a higher value for "Apache restart interval", like 1800 seconds for example. This won't really help for nginx as this setting only affects Apache but in my case it worked.

Then, once the migration is over, set the original value of the restart interval again.

 

[tkalfaoglu]

I remember I had a similar issue a couple of years ago. In my case, Apache/nginx was reloaded too often/quickly while Plesk was still busy building the configuration files for the migrated domains.

The migration continued and the errors disappeared after all domains have been migrated.
@tkalfaoglu You may want to try this: On the destination server go to "Apache & nginx settings" and set a higher value for "Apache restart interval", like 1800 seconds for example. This won't really help for nginx as this setting only affects Apache but in my case it worked.

Then, once the migration is over, set the original value of the restart interval again.

Many thanks! I am off to change that value

 

[Bitpalast]

Apache/nginx was reloaded too often/quickly while Plesk was still busy building the configuration files for the migrated domains.

Yes, that's the cause. And unfortunately, this does not only happen during migration, but it is a very broad and general issue that can also occur when impatient users play around with their configuration and happen to accidentally hit a moment when Apache or Nginx are restarting. It's also an issue when coincidentally SSL certificates are being generated or removed while configuration changes are being reloaded. Same issue, slightly different symptoms, but all the same root cause: Too many different scripts are able to tell the web server to reload or restart. That should all be queued instead, so that only a single process is able to control the web server.