Vladimir Sitnikov

Applications should log the exceptions properly. I think it is not an issue from the driver perspective. It would be great if you can link a guideline for drivers.

@AllanBoydCA , I mean that the driver itself does not log the exception. The following code in the issue description is out of pgjdbc's control: `System.out.println(">>> " + ex.getMessage() +...

Please provide the stacktrace or a steps to reproduce the issue. It is hard to analyze the issue otherwise.

I have not looked into this exact case, however, here's a way re-executing might be implemented: https://github.com/pgjdbc/pgjdbc/pull/451/files#diff-fb626514a44e1fd93551464de0ba369def2b0513a7232ce12d9c3040ea98d211R336 We could probably monitor "drop type" queries, and invalidate the relevant caches automatically....

``` 2023-12-06 11:21:34 FINEST QueryExecutorImpl sendParse FE=> Parse(stmt=null,query="drop type create_drop_struct",oids={}) 2023-12-06 11:21:34 FINEST QueryExecutorImpl sendBind FE=> Bind(stmt=null,portal=null) 2023-12-06 11:21:34 FINEST QueryExecutorImpl sendDescribePortal FE=> Describe(portal=null) 2023-12-06 11:21:34 FINEST QueryExecutorImpl sendExecute FE=>...

WDYT of something like https://github.com/apache/jmeter/pull/6478 instead?

>Or is this only possible when using the API and generate the test plan? Sure the idea is to support both UI and programmatic API. I just drafted the test...

>since the isEnabled() evaluates the JavaScript/Groovy code everytime? That is right. However, the default case is that "enabled" is either unset or set to `true`, so it should not be...

> it takes a LONG time to load because it is loading the entire table into RAM. As far as I remember, there's no way to use partial fetches in...

I must have been confused with conn.setAutoCommit(false); in the expected section