Hello Community,
I am currently experiencing an issue with my SuiteCRM setup and would appreciate your help.
SuiteCRM Version: 7.10.31
Production URL: productions.domain.com
Staging URL: staging.domain.com
Issue Description:
I have cloned my production environment to a staging environment using the following steps:
- Created a backup of the production database and files.
- Restored the backup on a different hosting account under the domain
staging.domain.com.
- Updated the
config.php file in the staging environment with the new database credentials, host_name, and site_url.
Config.php Updates:
$sugar_config['dbconfig']['db_host_name'] = 'staging_db_host';
$sugar_config['dbconfig']['db_user_name'] = 'staging_db_user';
$sugar_config['dbconfig']['db_password'] = 'staging_db_password';
$sugar_config['dbconfig']['db_name'] = 'staging_db_name';
$sugar_config['site_url'] = 'http://staging.domain.com';
$sugar_config['host_name'] = 'staging.domain.com';
Problem:
When I modify a field label using Studio in the staging environment, the same field label also changes in the production environment, and vice versa. This is unexpected as the two environments should be completely independent.
Steps Taken:
- Verified that the staging database is separate and uses different credentials from the production database.
- Ensured the
config.php file reflects the correct database information for each environment.
Questions:
- What could be causing changes in the staging environment to affect the production environment?
- Is there a specific configuration or step I might have missed during the cloning process to ensure the environments remain independent?
Any advice or suggestions on how to resolve this issue would be greatly appreciated.
Thank you in advance for your help!
Best regards
Maybe itās just a DNS issue, and youāre hitting the same IP address from both namesā¦
I am aware that both domains point to the same IP address, but they are set up in separate spaces on the server. Each domain has its own directory and separate database, and Iāve ensured that the config.php files in each environment reflect these separations correctly.
Iām sorry, maybe Iām just being dumb, but i that case I donāt understand how the web server directs each request to its own space?
Do you use a subfolder on the URL?
Apache would detect the hostname in the requested URL and go to the matching directory, itās called an Apache āVirtual Hostā.
Thanks Chris. I never did virtual hosts based on domain name, only based on subfolders or port numbers. But I guess it makes sense.
So the issue here would be to double-check that configuration, right? A web server issue, not a SuiteCRM issueā¦
Thanks, yes, Apache domain based VirtualHost-ing is how companies host many websites and other web services on one public IP address shared by all.
-
If you truly have separate hosting accounts, separate databases, then itās possible the previous domain name production.domain.tld hostname was copied in cached files from the production serverās Suite cache to the staging serverās Suite cache. To clear the cache on staging serverās Suite 7, go to Admin / Admin Tools / Repair / Quick Repair and Rebuild. Please post back with your results.
-
Double check your production and staging are not accidentally the same. If you can, share the Directory name, from the Apache VirtualHost configuration for each virtual server, here, they should be different directories. If you can, upload a small text file file to the Suite base directory of staging, it should not appear in the Suite base directory of production. Post back your results here.
I reconsidered this. Please open the .htaccess files which you copied over from the previous server. Copy paste them in your reply here to be sure. Do they contain the productions.mydomain.tld hostnames? You must change them to staging.mydomain.tld !
We are on a nginex server with directadmin. Each domain is configured to point to its own separate directory on the server.
Thank you for your response.
Regarding point one, I have performed the āQuick Repair and Rebuildā several times on both environments (production and staging). For added assurance, I also cleared the cache manually by deleting the contents of the cache folders via FTP on both environments, followed by running the āQuick Repair and Rebuildā again (both production and staging).
Regarding the server setup: it is a single multi-user/customerserver. Production and staging have two different accounts created via DirectAdmin, as if they were two separate customer of the same hosting provider. Each installation has its own base directory in public_html. If I upload a file to one siteās directory, it does not appear in the otherās directory. As proof of this, the two config.php files are different.
Any further insights or suggestions would be greatly appreciated.
Best regards,
I have already checked this as well: there are no references to the domain or local paths in the .htaccess files. Since they are in separate environments, both installations point to their respective roots, not to any subdirectories.
Itās crazy, but Iām even considering something that the browser might be caching locally on my computer. To rule this out, I have conducted tests using two different browsers in guest mode (without previous cache).
Try running the backup copy on your localhost (laptop or desktop) by restoring it on your local machine. Does it exhibit the same mysterious link to production?
I donāt know if this will work or not.
Got to admin ā System Settings ā Developer Mode:
Check if it is checked or not.
I checked and it is not checked.
Itās the last test I have left to do. Iāll let you know how it goes.
Could you share links & screenshots to show us?
Iām sorry, but I cannot share a link because the CRM contains private customer data. I can show you screenshots. What specifically would be helpful for you to see?
Hi Welcome to the forums,
The problem you are facing is rather strange, I canāt think of how the two could or are interplaying.
what web server are you using for fcgi? Iām guessing from reading the thread youāre using Apache.
One suggestion I can think of is to change the theme colour on one of the instances
(Iām not sure if 7.10 has the SuiteP theme, however a quick change to CSS regardless will also do),
You can then rule out any problems with DNS or vhost routing by expecting a certain UI colour.
The nature of your issue would suggest a problem with something in front of SuiteCRM in your setup rather than SuiteCRM itself, I would double check your vhost setup.
Thank you for your responses. It is indeed absurd. I am thinking of server-side issues. I already set different colors via CSS for staging a while ago to avoid confusion during testing. The configurations are definitely different; I have checked the config files and separate databases multiple times. If I create new fields, for example, I donāt find them duplicated on the other installation. The problem seems to be only with the labels of the fields that exist on both installations. My concern is that there may be other changes or settings that propagate between the two installations that I havenāt noticed yet.
As for the server, it is an Apache with Nginx in proxy, managed with DirectAdmin.
Here is an excerpt of the important parts from the end of the test site:
āsite_urlā => āhttps://test.domain.cloudā,
āhost_nameā => ātest.domain.cloudā,
ādb_host_nameā => ālocalhostā,
ādb_host_instanceā => āā,
ādb_user_nameā => ātest_',
ādb_passwordā => 'H6*******************ā,
ādb_nameā => ātest*****ā,
And here are the corresponding parts from the production site:
āsite_urlā => āhttps://prod.domain.cloudā,
āhost_nameā => āprod.domain.cloudā,
ādb_host_nameā => ālocalhostā,
ādb_host_instanceā => āā,
ādb_user_nameā => āprod_',
ādb_passwordā => 'k8*******************ā,
ādb_nameā => āprod*****ā,
Here is the screenshot of the two installations:
