Nowadays, there is little usage for raw blockchain export / import. In the past the p2p blockchain download was much slower. Back than blockchain.raw file was used to speed up download process of bootstrapping a node.
The tool imports bootstrap file `blockchain.raw` to your full node.
This could be useful if you want to decouple download from verification for any reason (like testing performance in isolation).
In this case, you may download the `blockchain.raw` file from [https://downloads.getmonero.org/blockchain.raw](https://downloads.getmonero.org/blockchain.raw).
Ideally though, you should use your own trusted `blockchain.raw` file that you [exported](../interacting/monero-blockchain-export-reference.md) earlier.
| `--log-level` | `0-4` with `0` being minimal logging and `4` being full tracing. Defaults to `0`. These are general presets and do not directly map to severity levels. For example, even with minimal `0`, you may see some most important `INFO` entries. Example: <br/>`./monero-blockchain-import --log-level=1`
| `--data-dir` | Full path to data directory. This is where the blockchain, log files, and p2p network memory are stored. For defaults and details see [data directory](../interacting/overview.md#data-directory).
| `--dangerous-unverified-import` | The safe default is to run verification (value `0`). You can enable `--dangerous-unverified-import` if you are importing from your own and trusted blockchain.raw (which we assume was already verified). The "dangerous" mode will greatly speed up the process.
| `--batch` | Whether to save to disk on an ongoing basis (the default, value `1`) or maybe do everything in RAM and save everything in the end (value `0`). No batching is only effective in combination with no verification (`--dangerous-unverified-import`). See also `--batch-size`.
| `--batch-size` | How often to save to disk expressed in number of blocks. By default save every `5000` blocks (when verifying) or every `20000` blocks (when not verifying). Big batches are faster but require more RAM.
| `--resume` | Resume from current height if output database already exists (the default, value `1`). Changing to `--resume=0` doesn't change much — existing blocks are skipped pretty quickly and the process is resumed anyway.