Hey!
I’ve finally come back to trying to uplift my instance from V4 to V5.
I tried a while back under this thread, but had too many issues and didnt have capacity at the time to resolve/test/trial on dev etc. Plus my prod PSU instance was in full flight with an IGA service (this is not currently being used so it’s an ideal time to upgrade).
I’ve started clean on a fresh server (windows), installed v5.4.3 msi as a service.
Then I’ve copied my repository codebase over, and replicated my app settings.
I got PSU up and running and OIDC working without issue.
Though my jobs are currently failing (parts of the scripts will work, but it fails when trying to issue any internal commands to the PSU management API such as Start-PSUApp).
Now I’ve read the previous thread, and noticed there’s changes and stricter mechanisms at play.
I updated my appsettings with the following:
“Api”: {
“SecurityModel”: “Permissive”,
“TrustCertificate”: “true”
}
After restarting the service and rerunning the job, I now get:
Status(StatusCode=“Unknown”, Detail=“Exception was thrown by handler. SqliteException: SQLite Error 19: ‘NOT NULL constraint failed: Apps.BaseUrl’.”)
For Reference im using a localhost cert in my appsettings:
“HTTPS”: {
“Url”: “https://*:443”,
“Certificate”:{
“Subject”: “localhost”,
“Store”: “My”,
“Location”: “LocalMachine”,
“AllowInvalid”: “true”
}
Also tried adding “URL”: “https://<serverhostname>/psu”
into the API block of my appsettings but still no change.
Also switched to Integrated mode instead of permissive.
In both cases above, I’m still getting the same ‘SQLite Error 19’ error
The SQL lite error isn’t related to the cmdlet communication and likely something else. Can you call something like Get-PSUCache without error? It seems like something is happening with Start-PSUApp. The SQL lite error seems to be referencing an attempt to create a new app in the internal database without a base URL.
I did Get-PSUCache -List and it threw 'the proxy tunnel request to proxy ‘http://proxyaddress’ failed with status code ‘403’.
(though i’ve been playing about with the settings, and realised i changed the cert from localhost to servername). Localhost was in the no_proxy env var but servername wasnt.
I had a lot of issues on v4 with my server sat behind a squid proxy, never fully resolved my problems but worked around them by only ever using the integrated environment when running scripts and setting the env vars for http_proxy/https_proxy and no_proxy.
Anyway, I just added the servername to no_proxy, rebooted tested again, and there was no error from issuing a Set-PSUCache and then a Get-PSUCache, I can see the output in the job log.
I’ve also reverted my appsettings to localhost too and tested and it doesnt throw an error. Though Start-PSUApp still throws the same error message.