Excited to give this a try. I tested almost every existing solution and as you said, nothing worked at scale. In the end I wrote my own tool specifically optim

Excited to give this a try. I tested almost every existing solution and as you said, nothing worked at scale. In the end I wrote my own tool specifically optimized for the use cases. It’s been rock solid for years so I know that better solutions are possible. But it has its limitations with little flexibility and I wouldn’t want to try an initial load of a table with 1B rows. It’s great to see someone with experience with large high throughput instances working on this.