jiaqizho commented on PR #900: URL: https://github.com/apache/cloudberry/pull/900#issuecomment-2655366207
> > My goal here is to do some "magic" for database users, not for us as database developers. > > As you mentioned, this does not benefit the kernel code in any way. On the contrary, it masks the problem and increases maintenance complexity. I stand by my point that this PR is not a good idea. > > Your example shows that ORCA is significantly slower when inserting data, but your conclusion generalizes to so-called simple queries. However, there are no examples or test data provided. For instance, what is the query plan for a statement like `select * from t`? This test seems simple, yet it is missing. Additionally, as I mentioned, what proportion of these so-called simple queries are actually `insert` statements? If the majority of the issues are caused by `insert`, we have even less reason to disable simple queries like `select`. > > I don’t mean to be rude, but while I find this requirement hardly worth discussing in terms of code, the code itself is quite sloppy. For example, the issue I raised about subqueries is not addressed in your code. I understand and agree with your point, but I also agree that we can provide some `simply and work` options (as the current PR does). As for maintenance concerns, my idea is that if the current `parse->rtable` can no longer be used as a IF condition (many case may lead to this), then we should think of other ways to solve the desired effect of the current PR. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
