How Blob Store Works Internally

How CB Server 5.0 handle blob store internally.
We have document json to store and blob data as well and our requirement is to store document data and corresponding blob data in atomic way but in different buckets (or tables) because every blob is linked with Json document and document should also have blob id.
We can not combine json and blob as it will increase the file size and will impact out DB write performance as we are working on top of large data set. Please suggest.

Divya Garg

What are you developing with and what platform are you on? You may want to select a category or better qualify the question with that, as there can be different approaches depending on mobile/server, platform, etc.

Also, note that your premise, storing things atomically but in a separate location, is challenging when you consider a lot of the possible failures and recoveries. There aren’t many systems that give you this kind of behavior across databases/buckets and they typically come at a high cost in performance, implementation complexity, etc. Usually it’s better if you can look at the specific use case to increase tolerance. There are bits of tech that exist for this, but they’re not broadly used.