Debian server with \po[ installed and worked well for some time. Cannot trace back what parameters changed. I stopped nginx for reducing possible side effects, using NS ony. I can open ]po[ login page using http://sub.domain.com:8000 , but after clicking "Connect" button the domain name in URL switches to servers hostname and the browser naturally shows "unable to connect". Where to search for a cause?
]project-open[ Version: 5.0.3.0.0
OS Platform: Linux
OS Version: 4.9.0-14-amd64
TCL Version: 8.5.18
Web Server Version: 4.99
Anonymous
Hi Antanas,
The server probably does a redirect to the original port.
Just make sure port 80 is also forwarded to the NaviServer...
Put Chrome in debugging mode (developer tools), and you'll see more.
Search in the database in the table apm_parameter_values for a value corresponding to your old server name.
Cheers
Frank
Dear Frank,
thank you for a quick feedback.
I should provide more details. Here are the sequential steps:
http://sub.domain.com:8000 – OK
...Entering false credentials...
http://sub.domain.com:8000/register/ – OK
...Entering valid credentials...
http://www1/intranet/ – can’t connect to the server ('www1' is the hostname of the server VM)
... manually correcting URL...
http://sub.domain.com:8000/intranet/ – OK
...Menu > Admin > Parameters...
http://sub.domain.com:8000/intranet/admin/parameters/ – OK
... Click on "acs-api-browser"...
http://sub.domain.com:8000/shared/parameters?package_id=129&return_url=%2fintranet%2fadmin%2fparameters%2f – OK
... Click Cancel...
http://www1/intranet/admin/parameters/ – can’t connect to the server
Last edit: Antanas 2020-12-04
Hi,
Please search for "www1" in apm_parameter_values via SQL and replace with "sub.domain.com"...
Cheers
Frank
Hi Frank,
not in this case:
The problem is reproducible.
1. On fresh ]po[ Debian install (following official recipe with all updates proposed through webUI) domain name remains intact while browsing the ]po[ site.
2. After
the starting domain name switches unexpectedly to server's hostname – wrong!
There are too many changed files (4335) to track down it fast.
Last edit: Antanas 2020-12-21
Hi Antanas,
There are two parameters controlling redirection.
Check:
Then check for the apm_parameter with the respective parameter_id.
Maybe also check ~/etc/config.tcl for the "hostname" parameter and set to the right URL.
Cheers
Frank
Hi Frank,
could you please moderate my previous posts send as Anonymous?
Sincerely,
Antanas
Hi, Frank,
I should repeat statements from my prior post which was sent accidentally as Anonymous a month ago.
Two servers are running PO instances: PO1 – production, PO2 – testing with demo data.
Both runs same PO version and both haven't the switching to wrong domain name (localhost) problem.
I'm backing up the PO1 data (Admin>Backup>New backup) and restoring it into PO2 server.
PO2 suddenly suffers from redirecting to wrong domain name.
What difference between two PO installs (some package? config value?) could impact on such unexpected behavior?
How to trace it down?