New Pricing Model - Up- or Downgrade?

[jammerdan]

@littleskunk

If you upload a file most pieces will end up on nodes close by because they win the long tail race.

Isn’t that something that has been even enforced through the new node selection algorithm that was introduced just a year ago?

@alpharabbit

The new node selection algorithm seems to prefer the closer nodes a lot and most customers are from the US.

@littleskunk

The current system would match the regional tier. If you upload a file most pieces will end up on nodes close by because they win the long tail race. Some pieces might end up on nodes further away.

I have always wondered about the piece distribution and the usefulness for users:

Distribution algorithm for node selection

I had uploaded 2 bigger files: File1 and File2 (Map creation takes a while) Both show that the distribution of pieces is mainly in the Europe and North America region and I am wondering if this is a desired outcome. In terms of redundancy and fast worldwide access, I would have expected a broader distribution. Both files seem to have not a single pieces stored in Australia very few in the Asia/Pacific region anyway. Almost none in Africa and very few again in South America. What is the reaso…

New operator in New Zealand. Fill rate seems dismal?

I always wondered the same: Distribution algorithm for node selection and I am still wondering. In the case above the file is small and maybe the new node selection algorithm has contributed to the outcome as well as optimizations like https://review.dev.storj.tools/c/storj/uplink/+/17523 where uploads get cancelled if they take longer than x-times the slowest of the fastest y uploads or even node exclusions like mentioned here: Automate node exclusions · Issue #7518 · storj/storj · GitHub. Bu…

Because close to the uploader is not always the preferred option for every use case.

I believe ideally uploads should be fast. This means minimum number of uploaded pieces, fast nodes, nearby, longtail cancellation and depending on upload resources either uploading in parallel with expansion or through gateway without. This should get a great and fast upload experience for users.

Same is true for downloads, however download location is not necessarily where the upload happened. It might even change after the data has been uploaded (just think of a camera to cloud upload in Turkey intended for the producer in Los Angeles that have to made available on the fly for post production improvements for a freshly contracted company in New Zealand.
So what I am trying to say is that at the time of uploading a user might not know what is the best distribution selection for his files. Close to himself at the time of uploading, close to one or more other locations or even varying locations.

My idea of an ideal solution would be that a user will always have the fastest possible upload and can choose on bucket level, what distribution he needs, more globally or more regionally. And for the best user experience this would mean that he can change this anytime and the network takes care of distributing pieces best suited for the intended coverage so that a user has to upload only once.