Hi team, We're currently considering using datahub...
# getting-started
b
Hi team, We're currently considering using datahub. I have played with the demo and searched through the docs (mostly the
Dataset
docs) but I can't find much regarding dataset versioning. 1. Is the same dataset mapped across different environments? E.g. if I have a dataset in dev and prod, are they automatically linked? Or are they treated as entirely different datasets because the URN is different? 2. Is there a versioning concept that allows to track datasets across multiple versions? It looks like schema changes are tracked, but can I tag those versions and use those tags to jump back in time to look at the definition of a table? 3. If it is possible to track versions: Is it possible to track a dataset when its table name changed? In major versions we might sometimes rename tables or move them to more fitting schemas. I've read somewhere that only non-breaking changes are supported, would a table rename be considered a breaking change?
1
m
1. They will be treat as different datasets 2. I believe versioning is a thing in dataset schema and it will be automatically tag when there is a change. You can check this demo dataset 3. I dont think table name change is tracked. It will be considered as a new one
m
Great questions @bumpy-architect-30714! Let's chat to understand the use-cases a bit better.