September 30, 2020 at 12:00 am
Comments posted to this topic are about the item The Value of Metadata
September 30, 2020 at 5:45 am
That is something that is really bothering me.
There seems so be so much useful metadata stored and available in SQL Server, but i dont even know where to start or what tutorial/guide/books to look for.
The only thing i am using is the INFORMATION_SCHEMA from time to time, but that's it.
I want to be the very best
Like no one ever was
September 30, 2020 at 8:02 am
That is something that is really bothering me.
There seems so be so much useful metadata stored and available in SQL Server, but i dont even know where to start or what tutorial/guide/books to look for.
The only thing i am using is the INFORMATION_SCHEMA from time to time, but that's it.
It's worth learning the system tables in SQL Server and how they hang together. All RDBMS have similar tables. This is Dr Codds rule 4. The database maintains its own catalogue
One thing I would advise people who build ETL pipelines is to make it data driven. Model the data mapping specs and get them imported into the DB so you have queryable metadata. One of the common things that I get asked is "what is the impact if we change 'x', what will it break"? Actually having the data lineage represented by a mapping spec as queryable data makes that a trivial task.
Note that Redgate SQLDoc scraped and publishes structural metadata and descriptions from SQL Server databases.
September 30, 2020 at 9:24 am
+1 on driving ETL pipelines through metadata where possible. I've implemented systems like that in a few places, and it can be more of an upfront cost but it makes everything so much easier to maintain.
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply