I’m facing a similar issue with a JDBC source connected to a MySQL database. It was working through Sept 15, but now fails to connect. Have you done similar investigations with JDBC connectors?
In our sandbox environment, the last successful connection to the database was on Sept. 15, the day that the JDK 17 upgrade went out. Since the JDK upgrade, the connector fails with error “The last packet sent successfully to the server was 0 milliseconds ago. The driver has not received any packets from the server”
The database owner says they see a “handshake error” on the server logs. I’ve tried upgrading and downgrading the jar file, but the results remain the same.
My main concern is that this will break production on Oct. 6, so we need a resolution or a way to prevent the JDK 17 upgrade before then. Our secondary concern is we are just beginning UAT on this implementation and the broken connector is a blocker to testing. Perhaps the case could even be argued that the testing is the bigger concern than the production connection right now.
We have a case open with support, but haven’t seen much activity on it, so I was reaching out here to see if anyone else in the community might have any ideas.
Hi All, we have identified the fix for JDBC sources. We will release the fix in sandbox on October 06 and postpose the production release. Will publish the new production release date in next week.
Seeing this message makes me wonder till what extend what this update to JDK17 tested before rolling it out in Sandbox? Of course it is impossible to test all the potential permutations in configurations that each client can have, but the error messages shared above on the JDBC source look to me like quick test cases that could’ve (should’ve?) been done prior to release, even in Sandbox.
Correct. We just pinned our CCG on Friday. Our sandbox’s authoritative source has been unavailable due to maintenance. We need extra time to perform a shakedown. You need to open a support case to pin CCG.
Today is the 6th. Has anyone received the new version in Production? If so, is there anything notable?
We opened a support case to Sailpoint, and they delivered a fix for our sandbox tenant yesterday. Our sqlloader, works now. The issue seems to be how the JDK17 handles https user/password.
From Sailpoint support:
”Oracle has introduced stricter security controls that prevent sensitive information such as usernames and passwords from being passed directly within the JDBC connection URL (e.g., jdbc:oracle:thin:user/password@host:port:sid)"his behavior was previously supported in earlier JDK versions, which is why the same configuration continues to work in their Prod environment (JDK 11).”
I hope this fix will be rolled also for others, and also then to the Production tenants. I guess that why the GoLive was postponed from 6Oct to “comming soon”
Hello Everyone, thank you for your interaction here, your valuable feedback matters a lot for us. I would like to share an update here, we are planning to release the changes to production on October 29, 2025 (CCG Version - 1164).
If there is any query or concern, please let us know here or raise a support ticket. Thanks!
I just want to clarify what you mean by custom VA-based connectors or connector-specific rules ?
Do we have to validate all our before \ after operation rules ? Does a standard Web Services connector (through VA) must be tested too following this upgrade ?
We have noticed that all the jdbc sources in our sandbox tenant are failing with the following error: ”[ InvalidConfigurationException ] [ Possible suggestions ] Verify that the driver details are correct/Missing required jar. [ Error details ] Error occurred before establishing the connection with Server. java.lang.ClassNotFoundException : test configuration”
We didn’t update our sources from long time and JAR files are present. We have even tried re-uploading the jar files as well, but it didn’t resolve the connectivity issue.
Can you please check if this has anything do with the latest CCG update?