Staged data with tokens demo notebook

[ognyanstoimenov]

Reference Issues/PRs

What does this implement or fix?

Add staged data notebook to the website.

Preview of the notebook: https://github.com/man-group/ArcticDB/blob/1ae8d5a2c6a75ce6a33b853b81401d0ab40af303/docs/mkdocs/docs/notebooks/ArcticDB_staged_data_with_tokens.ipynb

Any other comments?

Checklist

Checklist for code changes...

• Have you updated the relevant docstrings, documentation and copyright notice?
• Is this contribution tested against all ArcticDB's features?
• Do all exceptions introduced raise appropriate error messages?
• Are API changes highlighted in the PR description?
• Is the PR labelled as enhancement or bug so it appears in autogenerated release notes?

[jamesblackburn]

A few comments aiming to make this more user-centric.

Add:

• An introduction to Staging data, and description of the "Why" use-cases e.g. parallel loading of data. Then describe the high-level flow: stage data in parallel or sequentially, collect the tokens / cookies and finalise the staged data.

Remove:

• Old API: If we want the new way to be standard then no point referencing the old way

Change:

• New API -> API
• Document "stage result": i.e. every staged write returns a 'cookie/token' that is used for the finalisation stage.

Wy do we need this: -> Move to the introduction. Likely doesn't require code just description of how the API avoids pitfalls. Staging while finalising: similar just requires a comment rather than code.

How do you clean-up unstaged data in future runs? Should there be a clear or equivalent method to cleanup before running a large backfill.

[jamesblackburn]

Notebook looks good to me!