Qianqian

**Describe the bug** spiceai data connector returns lengthy error when dataset name is misconfigured **To Reproduce** use the following wrong configuration, which contains a wrong dataset from in spicepod.yaml ```yaml...

**Describe the bug** Using S3 connector and start the spice runtime with a wrongly configured endpoint. And then change the wrong endpoint to the correct one, the correct configuration doesn't...

- Close https://github.com/spiceai/spiceai/issues/1251

Enhancement: Improve spice runtime memory usage when accelerating with file-based accelerator in full refresh mode

1

## Goal-State/What/Result Large dataset won't get stuck in spice runtime memory and cause excessive memory usage / OOM of spice runtime when accelerating in full refresh mode with file-based accelerator....

DuckDB data connector should create a DuckDB file for user when .open parameter is not specified

2

- Scoped out from https://github.com/spiceai/spiceai/issues/1463

## 🗣 Description Enable parquet pushdown in Spice. This will fix the tpcds query 4, 11, 74, incorrectness issue where filter don't get applied in the non-pushdown mode. ## 🔨...

This reverts commit 641f692807c6db66a221ec08036c4d5398d3da2b. ## 🗣 Description ## 🔨 Related Issues ## 📄 Documentation Requirements ## 🤔 Concerns

TPCH Query 13 returns incorrect results for all SQL based connectors / accelerators. It's caused by unparser writing queries with duplicated filter.