Hello, folks!
Is it only me or is it a common issue that after installing APEX 18.1 when I try to reach my instance administration page (/apex_admin) or INTERNAL workspace, I always get the "ERR-7620 Could not determine workspace for application ()" exception - exactly like this, without the application ID specified?
Screenshot of the issue:
How did I get here:
- I had a bare-metal server with Oracle Database 11.2.0.2.0 XE, ORDS 17.4 and APEX 5.1.4 installed and worked properly altogether for my experiments.
- I got really excited about the APEX 18.1 general availability, so I decided to switch to it.
- I removed the previous version of APEX on my server totally by running the apxremov.sql script from its working directory (remove was successful). Then deleted the files of previous version of APEX totally.
- Downloaded the package Oracle APEX 18.1.0 - English language only (http://download.oracle.com/otn/java/appexpress/apex_18.1_en.zip) to my server.
- Unpacked the archive to a working directory.
- Installed the brand new version of APEX using apexins.sql script and the official documentation (installation process was successful and without errors)
- Reconfigured ORDS using apex_rest_config.sql script
- Reconfigured APEX instance admin user using apxchpwd.sql script
- Updated the APEX static files by replacing the old ones with the new in my environment
- Restarted the services (httpd and tomcat in my case)
- Opened my browser and typed the APEX URL, it worked, but I cannot access any workspace now (including INTERNAL) - when I press the LOGIN button, I constntly get the same exception.
I also found a useful script for support needs (%APEX_HOME%/utilities/support/apex_verify.sql) which generates a lot of useful supporting information (though I haven't seen anything new for me there): http://bit.ly/apex_support_info
Updated
What I already tried:
- Reinstalled APEX 18.1 for the second time (removed and installed again using the same steps) - didn't help.
- Looked through APEX installation logs - no errors found.
- Looked through ORDS installation logs - no errors found.
- Revalidate ORDS and looked through validation logs (no errors found), then restarted tomcat service - didn't help.
- Reinstalled ORDS totally (using uninstall and install commands of ORDS.war) - didn't help.
- Looked through various Linux log files - nothing useful was found.
- Downgraded to APEX 5.1.4 (removed 18.1 using the apxremov.sql script from its working directory, then went to the 5.1.4 working directory and repeated the installation process from there) - and the problem persisted after this step - I was really surprised.
- Reinstalled APEX 18.1 again - didn't help.
- Created a fresh virtual machine with the same OS (CentOS 7.4), installed and configured all the software from scratch - and IT WORKED! So, the problem is with that exact environment and that exact server - though I don't understand how it's different.
- Looked through the report generated by the report_instance_configuration.sql script - no deviations.
- Looked thtough the report generated by %APEX_HOME%/utilities/debug/activity.sql script. Nothing really useful.
- Searched for the "ERR-7620" string in file contents on my file system (maybe I'm missing some log files?). Nothing really useful.
- Reinstalled ORDS 18.1 under a different directory. Didn't help.
- Installed a different browser (Opera this time) on my desktop computer to be sure it is not a browser-specific thing and not cache. The issue persisted.
Any ideas where to look next?