influxdb/influxdb3_server/tests/snapshots
praveen-influx 277e153a44
feat: upgrade to hyper 1 (#26655)
* feat: upgrade to hyper 1

- use `hyper_util` for `TokioIo` and `ConnectionBuilder` from it
- remove `hybrid` service running grpc/http on same port to new
  `UnifiedService`, uses less generics
- swap `hyper_util::client::legacy::Client` for `hyper::Client`
- TLS changes, set ALPN protocol
- Test code changes, instead of `hyper::{StatusCode, Method}` use `reqwest::{StatusCode, Method}`
- rustls initialization (crypto provider) needs to be done explicitly now
- graceful shutdown + tidy ups
- move tokio-rustls to root Cargo.toml

helps: https://github.com/influxdata/influxdb_pro/issues/1076

* feat: upgrade all non-hyper libraries

- update arrow/datafusion/object_store/parquet related dependencies to align with `iox` (or `influxdb3_core`)
- move of `datafusion::physical_plan::memory::MemoryExec`, the actual alternative is to use `MemorySourceConfig` and `DataSourceExec` directly
- move from `use parquet_file::storage::ParquetExecInput;` to `use parquet_file::storage::DataSourceExecInput;`
- object_store life time requirement changes, mostly switch to `'static`
- object_store crate deprecating `PutMultiPartOpts` in favour of `PutMultiPartOptions`
- `Range<usize>` to `Range<u64>` move in object_store. Most of them are updates to method signatures in impls but the parquet_cache one which needed bit more attention

closes: https://github.com/influxdata/influxdb_pro/issues/1076

* refactor: address feedback
2025-08-12 20:01:23 +01:00
..
lib__deduplicate_rows_in_write_buffer_both_memory_and_parquet.snap feat: upgrade to hyper 1 (#26655) 2025-08-12 20:01:23 +01:00
lib__deduplicate_rows_in_write_buffer_memory.snap test: add integration tests to influxdb3_server (#26474) 2025-05-28 21:39:40 -04:00
lib__deduplicate_rows_in_write_buffer_parquet.snap feat: upgrade to hyper 1 (#26655) 2025-08-12 20:01:23 +01:00