Log errors when doing version compatibility check for Cassandra and SQL

[simvlad]

What changed?

Changes version compability checks to pass the real logger, so that the error messages would be seeing in the log

Why?

Before, the error was generic and there were no logs indicating the real error (see #7939):

sql schema version compatibility check failed: unable to read DB schema version keyspace/database: temporal error: no usable database connection found

With this change, the log would contain the actual error:

{"level":"error","ts":"2025-07-25T15:30:49.660-0700","msg":"sql handle: unable to refresh database connection pool","error":"pq: no pg_hba.conf entry for host \"127.0.0.1\", ...

The alternative solution would be to return the error, but the design of common/persistence/sql/sqlplugin/db_handle.go suggests we hold the abstraction boundaries, by returning generic errors and logging specific errors.

How did you test it?

• built
• run locally and tested manually
• covered by existing tests

[yycptt]

the design of common/persistence/sql/sqlplugin/db_handle.go suggests we hold the abstraction boundaries, by returning generic errors and logging specific errors.

I was thinking only return the error on NewDatabaseHandle which eagerly sets up the connection. Didn't realize we are using no-op logger today and not even logging the real error. This approach LGTM.