Lin Ma

Decouple the DataTable pointer from the plan objects

3

We now have the `storage::DataTable` pointer inside many plan objects, such as `AbstractScanPlan`, `AnalyzePlan`, `DeletePlan`, `InsertPlan`, and `UpdatePlan`. This prevents us from shipping the plan objects around or using the...

We are now leveraging the settings manager to invoke self-driving related functionalities (e.g., `pilot_planning`, `train_forecast_model`, and `clear_query_cache`). That's not what exactly the setting manager is designed for. Instead, we may...

Error catching for internal tasks (queries)

1

We added support for internal tasks in #1502, but we don't catch the errors when the internal tasks fail. Instead, we just print our error log messages from the command...

Should we remove the gbenchmark dependency from the mini runners?

2

We implement the mini (operating unit) runners on top of the gbenchmark framework. We use the setup/teardown, benchmark filter, and the argument generation features of gbenchmark. But we're not using...

The way that the mini-runners work right now is to first create all the tables required to enumerate input features and then and execute the runners for every OU one-by-one...

I've tested YCSB and TPCC benchmark on the asynchronous eviction branch (https://github.com/zheguang/h-store/commits/master). I'm always hitting the "failed to update tuple to new address" error. Script and example log for YCSB...