Rohit Nayak

Squashed all commits since it was becoming tougher to fix conflicts each time with a large number of commits..

It looks like, today based on the documentation and test cases, Vitess expects reference tables to have the same name on the sharded keyspace as the unsharded. I was able...

OK, then there is a regression. We had made a few changes around reference tables in v17, so that might have caused it or possibly some other refactoring in vtgate...

I am not able to repro this in v19. I ran the end to end test `TestReferenceRouting` in `go/test/endtoend/vtgate/queries/reference`. I commented out the teardown `// defer clusterInstance.Teardown()` . Based on...

> I can work around it, but it seems to me it's important to fix, because if I add extra reference tables in the future without adding to the vschema,...

> One extra problem though: I get a panic in vtgate when I use unspecified mode in my application now. I will discuss/report that separately. I see you logged the...

We need to wait till we reach a low replication lag before we can route the reference table to the materialized keyspace in the VSchema. Otherwise we will be querying...

@mattlord: `TestSchemaChange/revertible/expanded:_enum` is failing

Since there are no reviews yet, closing this in favor of https://github.com/vitessio/vitess/pull/15366 which aims to port all the tests in `vstreamer_test.go`.

> Do we have end to end tests for unmanaged tablets @rohit-nayak-ps? If we do, we should add this flag there to make sure that the validations are correct. We...