vegetableysm

What do these changes do? ------------------------- As tittled. Related issue number -------------------- Fixes #1733

Describe your problem --------------------- Avoid pull all the cache block from remote. Do not add all block objects to cache object as members. Additional context ------------------ Add any other context...

Describe your problem --------------------- When spark insert data into hive tables, the stage will be blocked if the spark and hive-metastore in different container. If they are in same container,...

### What is the problem: Current hive modules can work on multi node [pr-1624](https://github.com/v6d-io/v6d/pull/1624) but queries are distributed randomly across nodes. ### The behaviour that you expect to work: Use...

Describe your problem --------------------- - [ ] Choose a test framework - [ ] Test hive module performance and compare with competitors - [ ] Optimize performance

Describe your problem --------------------- Currently, hive can create and query tables on the vineyard, but there is no way to retrieve and query arrow tables created on the vineyard by...

Describe your problem --------------------- The data for a hive table can be stored in different files under the same folder. However, the hive tables currently stored on the vineyard are...

Analyze the query data in the actual scene, and provide a basis for the subsequent parameter setting.

1

Describe your problem --------------------- As tittled.

Vineyard memory was occasionally incompletely recycled when the llm test exited(blob storage)

3

Describe your problem --------------------- As tittled. Currently disable vineyard memory usage checks of llm test provisionally.(Fix in the future)

Create llm cache block builder only when the user requests an update.

3

Describe your problem --------------------- Currently, we will fetch and get a block object if user request a cache block(update or query) and create a builder from this block object. Then...